home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
CU Amiga Super CD-ROM 13
/
CU Amiga Magazine's Super CD-ROM 13 (1997)(EMAP Images)(GB)(Track 1 of 2)[!][issue 1997-08].iso
/
CUCD
/
Graphics
/
irit70
/
bin
/
irit.hlp
< prev
next >
Wrap
Text File
|
1997-05-01
|
326KB
|
8,984 lines
IRIT Version 7.0 User's Manual
A Solid modeling Program
Copyright (C) 1989, 1990-1997 Gershon Elber
EMail: gershon@cs.technion.ac.il
Join IRIT mailing list: gershon@cs.technion.ac.il
Mailing list: irit-mail@cs.technion.ac.il
Bug reports: irit-bugs@cs.technion.ac.il
WWW Page: http://www.cs.technion.ac.il/~irit
Introduction
IRIT is a solid modeler developed for educational purposes. Although
small, it is now powerful enough to create quite complex scenes.
IRIT started as a polygonal solid modeler and was originally developed
on an IBM PC under MSDOS. Version 2.0 was also ported to X11 and version 3.0
to SGI 4D systems. Version 3.0 also includes quite a few free form curves
and surfaces tools. See the UPDATE.NEW file for more detailed update
information. In Version 4.0, the display devices were enhanced, freeform
curves and surfaces have further support, functions can be defined, and
numerous improvement and optimizations are added.
$
COPYRIGHTS
BECAUSE IRIT AND ITS SUPPORTING TOOLS AS DOCUMENTED IN THIS DOCUMENT
ARE LICENSED FREE OF CHARGE, I PROVIDE ABSOLUTELY NO WARRANTY, TO THE EXTENT
PERMITTED BY APPLICABLE STATE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING, I
GERSHON ELBER PROVIDE THE IRIT PROGRAM AND ITS SUPPORTING TOOLS "AS IS"
WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
PARTICULAR PURPOSE.
THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THESE PROGRAMS IS WITH
YOU. SHOULD THE IRIT PROGRAMS PROVE DEFECTIVE, YOU ASSUME THE COST OF
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW WILL GERSHON ELBER,
BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY LOST PROFITS, LOST MONIES,
OR OTHER SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
USE OR INABILITY TO USE (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR A
FAILURE OF THE PROGRAMS TO OPERATE WITH PROGRAMS NOT DISTRIBUTED BY GERSHON
ELBER) THE PROGRAMS, EVEN IF YOU HAVE BEEN ADVISED OF THE POSSIBILITY OF
SUCH DAMAGES, OR FOR ANY CLAIM BY ANY OTHER PARTY.
IRIT is a freeware solid modeler. It is not public domain since I
hold copyrights on it. However, unless you are to sell or attempt to make
money from any part of this code and/or any model you made with this solid
modeler, you are free to make anything you want with it.
IRIT can be compiled and executed on numerous Unix systems as well
as OS2, Windows NT/95 and AmigaDOS. However, beware the MSDOS support is
fading away.
You are not obligated to me or to anyone else in any way by using IRIT.
You are encouraged to share any model you made with it, but the models
you made with it are yours, and you have no obligation to share them.
You can use this program and/or any model created with it for non
commercial and non profit purposes only. An acknowledgement on the way the
models were created would be nice but is not required.
$
Setup
The IRIT program reads a file called irit.cfg each time it is executed.
This file configures the system. It is a regular text file with comments, so
you can edit it and properly modify it for your environment.
This file is being searched for in the directory specified by the
IRIT_PATH environment variable.
For example 'setenv IRIT_PATH /u/gershon/irit/bin/'.
Note IRIT_PATH must terminate with '/'. If the variable is not set only
the current directory is being searched for irit.cfg.
In addition, if it exists, a file by the name of iritinit.irt will be
automatically executed before any other '.irt' file. This file may contain
any IRIT command. It is the proper place to put your predefined
functions and procedures if you have some.
This file will be searched much the same way IRIT.CFG is. The
name of this initialization file may be changed by setting the StartFile
entry in the configuration file.
This file is far more important starting at version 4.0, because of the new
function and procedure definition that has been added, and which is used
to emulate BEEP, VIEW, and INTERACT for example.
The solid modeler can be executed in text mode (see the .cfg and the -t
flag below) on virtually any system with a C compiler.
Under all systems the following environment variables must be set
and updated:
path Add to path the directory where IRIT's binaries are.
IRIT_PATH Directory with config., help and IRIT's binary files.
IRIT_DISPLAY The graphics driver program/options. Must be in path.
IRIT_BIN_IPC If set, uses binary Inter Process Communication.
For example,
set path = (path /u/gershon/irit/bin)
setenv IRIT_PATH /u/gershon/irit/bin/
setenv IRIT_DISPLAY "xgldrvs -s-"
setenv IRIT_BIN_IPC 1
to set /u/gershon/irit/bin as the binary directory and to use the sgi's
gl driver. If IRIT_DISPLAY is not set, the server (i.e., the IRIT
program) will prompt and wait for you to run a client (i.e., a display
driver). if IRIT_PATH is not set, none of the configuration files, nor
the help file will be found.
If IRIT_BIN_IPC is not set, text based IPC is used, which is far
slower. No real reason not to use IRIT_BIN_IPC, unless it does not
work for you.
In addition, the following optional environment variables may be set.
IRIT_MALLOC If set, apply dynamic memory consistency testing.
Programs will execute much slower in this mode.
IRIT_MALLOC_PTR Set to a pointer address and the program will scream
once this pointer is allocated, if IRIT_MALLOC is set.
IRIT_NO_SIGNALS If set, no signals are caught by IRIT.
IRIT_SERVER_HOST Internet Name of IRIT server (used by graphics driver).
IRIT_SERVER_PORT Used internally to the TCP socket number. Should not
be set by users.
LD_LIBRARY_PATH If shared libraries are created, this variable must be
updated to point to the shared libraries' directory.
For example,
setenv IRIT_MALLOC 1
setenv IRIT_MALLOC_PTR 1234567890
setenv IRIT_NO_SIGNALS 1
setenv IRIT_SERVER_HOST irit.cs.technion.ac.il
IRIT_MALLOC is useful for programmers, or when reporting a memory
fatal error occurrence. This variable, when set as a non zero value
will activate the following (hexadecimal bit settings with any combination
of the following):
0x01 A test for overwriting before the dynamic memory
allocated or immediately after it. Cheap in time.
0x02 Savings of all allocated objects in a table for the
detection of freeing unallocated objects and consistency
of the entire dynamic memory. Expensive in time.
0x04 Zeros every freed object, once it is freed.
IRIT_NO_SIGNALS is also useful for debugging when contorl-C is used
within a debugger. The IRIT_SERVER_HOST/PORT controls the server/client
(IRIT/Display device) communication.
IRIT_SERVER_HOST and IRIT_SERVER_PORT are used in the unix and
Window NT ports of IRIT.
See the section on the graphics drivers for more details.
A session can be logged into a file as set via LogFile in the configuration
file. See also the LOGFILE command.
The following command line options are available:
IRIT [-t] [-z] [file.irt]
-t Puts IRIT into text mode. No graphics will be displayed and
the display commands will be ignored. Useful when one needs to
execute an irt file to create data on a tty device...
-z Prints usage message and current configuration/version
information.
file.irt A file to invoke directly instead of waiting to input from
stdin.
Under OS2 the IRIT_DISPLAY environment variable must be set (if set) to
os2drvs.exe without any option (-s- will be passed automatically).
os2drvs.exe must be in a directory that is in the PATH environment
variable. IRIT_BIN_IPC can be used to signal binary IPC which is faster.
Here is a complete example:
set IRIT_PATH=c:\irit\bin\
set IRIT_DISPLAY=os2drvs -s-
set IRIT_BIN_IPC=1
assuming the directory specified by IRIT_PATH holds the executables of
IRIT and is in PATH.
If IRIT_BIN_IPC is not set, text based IPC is used which is far
slower. No real reason not to use IRIT_BIN_IPC unless it does not
work for you.
The OS2 executables are typically built using the EMX port of gnu C
compiler. The distribution of the executables does not include the run
time library of EMX and attempt to run IRIT will fail with error message
like "File EMX does not exist". You will need to get the run time from
ftp to ftp-os2.nmsu.edu (aliased also as hobbes.NMSU.Edu)
cd to os2/unix/emx09c (or a newer version number/level)
get emxrt.zip and place its dlls in a place they would be found.
The NT port uses sockets and is, in this respect, similar to the unix port.
The envirnoment variables IRIT_DISPLAY, IRIT_SERVER_HOST,
IRIT_BIN_IPC should all be set in a similar way to the Unix specific setup.
As a direct result, the server (IRIT) and the display device may be running
on different hosts. For example the server might be running on an NT system
while the display device will be running on an SGI4D exploiting the graphic's
hardware capabilities. Here is a complete example:
set IRIT_PATH=c:\irit\bin\
set IRIT_DISPLAY=wntgdrvs -s-
set IRIT_BIN_IPC=1
Under UNIX using X11 (x11drvs driver) add the following options to
your .Xdefaults. Most are self explanatory. The Trans attributes control
the transformation window, while the View attributes control the view window.
SubWin attributes control the subwindows within the Transformation window.
#if COLOR
irit*Trans*BackGround: NavyBlue
irit*Trans*BorderColor: Red
irit*Trans*BorderWidth: 3
irit*Trans*TextColor: Yellow
irit*Trans*SubWin*BackGround: DarkGreen
irit*Trans*SubWin*BorderColor: Magenta
irit*Trans*Geometry: =150x500+500+0
irit*Trans*CursorColor: Green
irit*View*BackGround: NavyBlue
irit*View*BorderColor: Red
irit*View*BorderWidth: 3
irit*View*Geometry: =500x500+0+0
irit*View*CursorColor: Red
irit*MaxColors: 15
#else
irit*Trans*Geometry: =150x500+500+0
irit*Trans*BackGround: Black
irit*View*Geometry: =500x500+0+0
irit*View*BackGround: Black
irit*MaxColors: 1
#endif
$
First Usage
Commands to IRIT are entered using a textual interface, usually
from the same window the program was executed from.
Some important commands to begin with are,
1. include("file.irt"); - will execute the commands in file.irt. Note
include can be recursive up to 10 levels. To execute the demo
(demo.irt) simply type 'include("demo.irt");'. Another way to run
the demo is by typing demo(); which is a predefined procedure defined
in iritinit.irt.
2. help(""); - will print all available commands and how to get help on
them. A file called irit.hlp will be searched as irit.cfg is
being searched (see above), to provide the help.
3. exit(); - close everything and exit IRIT.
Most operators are overloaded. This means that you can multiply
two scalars (numbers), or two vectors, or even two matrices, with the same
multiplication operator (*). To get the on-line help on the
operator '*' type 'help("*");'
The best way to learn this program (like any other program...) is by
trying it. Print the manual and study each of the commands available.
Study the demo programs (*.irt) provided as well.
The "best" mode to use irit is via the emacs editor. With this distribution
an emacs mode for irit files (irt postfix) is provided (irit.el). Make your
.emacs load this file automatically. Loading file.irt will switch emacs into
Irit mode that supports the following three keystrokes:
Meta-E Executes the current line
Meta-R Executes the current Region (Between Cursor and Mark)
Meta-S Executes a single line from input buffer
The first time one of the above keystrokes is hit, emacs will fork an Irit
process so that Irit's stdin is controlled via the above commands.
This emacs mode was tested under various unix environments, under OS2
2.x/3.x, and under Windows NT 3.x/4.x and Windows 95.
$
Data Types
These are the Data Types recognized by the solid modeler. They are also
used to define the calling sequences of the different functions below:
ConstantType Scalar real type that cannot be modified.
NumericType Scalar real type.
VectorType 3D real type vector.
PointType 3D real type point.
CtlPtType Control point of a freeform curve or surface.
MatrixType 4 by 4 matrix (homogeneous transformation matrix).
PolygonType Object consists of polygons.
PolylineType Object consists of polylines.
CurveType Object consists of curves.
SurfaceType Object consists of surfaces.
TrimSrfType Object consists of trimmed surfaces.
TriSrfType Object consists of triangular surfaces
TrivarType Object consists of trivariate functions.
GeometricType One of Polygon/lineType, CurveType, SurfaceType,
TrimSrfType, TrivarType, TriSrfType.
InstanceType Object with a GeometryType and a Transformation.
GeometricTreeType A list of GeometricTypes or GeometricTreeTypes.
StringType Sequence of chars within double quotes - "A string".
Current implementation is limited to 80 chars.
AnyType Any of the above.
ListType List of (any of the above type) objects. List
size is dynamically increased, as needed.
Although points and vectors are not the same, IRIT does not
destinguish between them, most of the time. This might change in the future.
$
Commands
These are all the commands and operators supported by the IRIT solid
modeler:
+ CHDIR CZEROS LN SBEZIER SYMBSUM
- CINFLECT DSTPTLN LOAD SBISECTOR SYSTEM
* CINTERP DSTPTPLN LOFFSET SBSPLINE TAN
/ CIRCLE DSTLNLN LOG SCALE TBEZIER
^ CIRCPOLY ERROR LOGFILE SCRVTR TBSPLINE
= CLNTCLOSE EVOLUTE MERGEPOLY SDERIVE TCRVTR
== CLNTEXEC EXIT MESHSIZE SDIVIDE TDERIVE
!= CLNTREAD EXP MOFFSET SEDITPT TEDITPT
< CLNTWRITE EXTRUDE MOMENT SEVAL TEVAL
> CMESH FFCOMPAT MRCHCUBE SFOCAL TEXTGEOM
<= CMORPH FFCTLPTS MSLEEP SFROMCRVS TFROMSRFS
>= CMULTIRES FFEXTREME NIL SGAUSS TIME
ABS CNORMAL FFKNTVEC NTH SIN TINTERP
ACOS COERCE FFMATCH OFFSET SINTERP THISOBJ
ADAPISO COLOR FFMERGE ORTHOTOMC SIZEOF TMORPH
ARC COMMENT FFMSIZE PAUSE SMEANSQR TORUS
AREA COMPOSE FFORDER PCIRCLE SMERGE TRAISE
ASIN CON2 FFPOLES PDECIMATE SMORPH TRANS
ATAN CONE FFPTDIST PDOMAIN SNOC TREFINE
ATAN2 CONTOUR FFPTTYPE PLN3PTS SNORMAL TREGION
ATTRIB CONVEX FFSPLIT POLY SNRMLSRF TREPARAM
AOFFSET COORD FLOOR POWER SPHERE TRIANGL
AWIDTH COS FMOD PRINTF SQRT TRIMSRF
BBOX COVERISO FOR PRISA SRAISE TSBEZIER
BOOLONE COVERPT FREE PROCEDURE SREFINE TSBSPLINE
BOOLSUM CPOLY FUNCTION PT3BARY SREGION TSDERIVE
BOX CRAISE GBOX PTHMSPR SREPARAM TSEVAL
BSP2BZR CREFINE GETATTR PTLNPLN SRINTER TSNORMAL
BZR2BSP CREGION GETLINE PTPTLN STANGENT TVLOAD,
CBEZIER CREPARAM GPOLYGON PTSLNLN STRIMSRF VARLIST
CBISECTOR CROSSEC GPOLYLINE RANDOM STRIVAR VECTOR
CBSPLINE CRV2TANS HELP RMATTR SURFPREV VIEW
CCINTER CRVLNDST HERMITE ROTVEC SURFREV VIEWOBJ
CCRVTR CRVPTDST HOMOMAT ROTX SVISIBLE VOLUME
CDERIVE CRVPTTAN IF ROTY SWEEPSRF WHILE
CDIVIDE CSURFACE INCLUDE ROTZ SWPSCLSRF
CEDITPT CTANGENT INSTANCE ROTZ2V, SYMBCPROD
CENVOFF CTLPT INTERACT ROTZ2V2, SYMBDIFF
CEVAL CTRIMSRF IRITSTATE RULEDSRF SYMBDPROD
CEXTREMES CYLIN LIST SAVE SYMBPROD
$
Functions and Variables
Functions that return a NumericType:
ABS COS EXP POWER THISOBJ
ACOS CLNTEXEC FLOOR RANDOM VOLUME
AREA CPOLY FMOD SIN
ASIN DSTPTLN LN SIZEOF
ATAN DSTPTPLN LOG SQRT
ATAN2 DSTLNLN MESHSIZE TAN
Functions that return a GeometricType:
ADAPISO CON2 FFORDER PTSLNLN SWEEPSRF
ARC CONE FFPOLES RULEDSRF SWPSCLSRF
AOFFSET CONTOUR FFPTDIST SBEZIER SYMBCPROD
BBOX CONVEX FFPTTYPE SBISECTOR SYMBDIFF
BOOLONE COORD FFSPLIT SBSPLINE SYMBDPROD
BOOLSUM COVERISO GBOX SCRVTR SYMBPROD
BOX COVERPT GETATTR SDERIVE SYMBSUM
BSP2BZR CRAISE GETLINE SDIVIDE TBEZIER
BZR2BSP CREFINE GPOLYGON SEDITPT TBSPLINE
CBEZIER CREGION GPOLYLINE SEVAL TCRVTR
CBISECTOR CREPARAM HERMITE SFOCAL TDERIVE
CBSPLINE CROSSEC INSTANCE SFROMCRVS TEDITPT
CCINTER CRV2TANS LOFFSET SGAUSS TEVAL
CCRVTR CRVLNDST MERGEPOLY SINTERP TEXTGEOM
CDERIVE CRVPTDST MOFFSET SMEANSQR TFROMSRFS
CDIVIDE CRVPTTAN MOMENT SMERGE TINTERP
CEDITPT CSURFACE MRCHCUBE SMORPH TMORPH
CENVOFF CTANGENT NIL SNORMAL TORUS
CEVAL CTRIMSRF OFFSET SNRMLSRF TRAISE
CEXTREMES CTLPT ORTHOTOMC SPHERE TREFINE
CINFLECT CYLIN PCIRCLE SRAISE TREGION
CINTERP CZEROS PDECIMATE SREFINE TREPARAM
CIRCLE EVOLUTE PDOMAIN SREGION TRIANGL
CIRCPOLY EXTRUDE PLN3PTS SREPARAM TRIMSRF
CLNTREAD FFCOMPAT POLY SRINTER TSBEZIER
CMESH FFCTLPTS PRISA STANGENT TSBSPLINE
CMORPH FFEXTREME PROCEDURE STRIMSRF TSDERIVE
CMULTIRES FFKNTVEC PT3BARY STRIVAR TSEVAL
CNORMAL FFMATCH PTHMSPR SURFPREV TSNORMAL
COERCE FFMERGE PTLNPLN SURFREV TVLOAD
COMPOSE FFMSIZE PTPTLN SVISIBLE
Functions that create linear transformation matrices:
HOMOMAT ROTX ROTZ ROTZ2V2 TRANS
ROTVEC ROTY ROTZ2V SCALE
Miscellaneous functions:
ATTRIB ERROR INCLUDE NTH SYSTEM
AWIDTH EXIT INTERACT PAUSE TIME
CHDIR FOR IRITSTATE PRINTF VARLIST
CLNTCLOSE FREE LIST PROCEDURE VECTOR
CLNTWRITE FUNCTION LOAD RMATTR VIEW
COLOR HELP LOGFILE SAVE VIEWOBJ
COMMENT IF MSLEEP SNOC WHILE
Variables that are predefined in the system:
AXES MACHINE POLY_APPROX_TOL VIEW_MAT
DRAWCTLPT POLY_APPROX_OPT PRSP_MAT
FLAT4PLY POLY_APPROX_UV RESOLUTION
Constants that are predefined in the system:
AMIGA E3 MAGENTA PARAM_CENTRIP SURFACE_TYPE
APOLLO E4 MATRIX_TYPE PARAM_CHORD SUN
BLACK E5 MSDOS PARAM_UNIFORM TRIMSRF_TYPE
BLUE FALSE NUMERIC_TYPE PI TRIVAR_TYPE
COL GREEN OFF PLANE_TYPE TRUE
CTLPT_TYPE HP ON POINT_TYPE UNDEF_TYPE
CURVE_TYPE IBMOS2 P1 POLY_TYPE UNIX
CYAN IBMNT P2 RED VECTOR_TYPE
DEPTH KV_FLOAT P3 ROW WHITE
E1 KV_OPEN P4 SGI YELLOW
E2 LIST_TYPE P5 STRING_TYPE
$
Language Description
The front end of the IRIT solid modeler is an infix parser that
mimics some of the C language behavior. The infix operators that are supported
are plus (+), minus (-), multiply (*), divide (/), and power (^), for
numeric operators, with the same precedence as in C.
However, unlike the C language, these operators are overloaded,
or different action is taken, based upon the different operands.
This means that one can write '1 + 2', in which the plus sign denotes a
numeric addition, or one can write 'PolyObj1 + PolyObj2', in which case the
plus sign denotes the Boolean operation of a union between two geometric
objects.
The exact way each operator is overloaded is defined below.
In this environment, reals, integers, and even Booleans, are all represented
as real types. Data are automatically promoted as necessary.
For example, the constants TRUE and FALSE are defined as 1.0 and 0.0
respectively.
Each expression is terminated by a semicolon. An expression can be as
simple as 'a;' which prints the value of variable a, or as complex as:
for ( t = 1.1, 0.1, 1.9,
cb1 = csurface( sb, COL, t ):
color( cb1, green ):
snoc( cb1, cb_all )
);
While an expression is terminated with a semicolon, a colon is used to
terminate mini-expressions within an expression.
Once a complete expression is read in (i.e., a semicolon is detected)
and parsed correctly (i.e. no syntax errors are found), it is executed.
Before each operator or a function is executed, parameter type matching
tests are made to make sure the operator can be applied to these
operand(s), or that the function gets the correct set of arguments.
The parser is totally case insensitive, so Obj, obj, and OBJ will refer
to the same object, while MergePoly, MERGEPOLY, and mergePoly will refer
to the same function.
Objects (Variables if you prefer) need not be declared. Simply use them
when you need them. Object names may be any alpha-numeric (and underscore)
string of at most 30 characters. By assigning to an old object, the old
object will be automatically deleted and if necessary its type will be
modified on the fly.
Example:
V = sin( 45 * pi / 180.0 );
V = V * vector( 1, 2, 3 );
V = V * rotx( 90 );
V = V * V;
will assign to V a NumericType equal to the sine of 45 degrees, the VectorType
( 1, 2, 3 ) scaled by the sine of 45, rotate that vector around the X axis
by 90 degrees, and finally a NumericType which is the dot (inner) product of
V with itself.
The parser will read from stdin, unless a file is specified on the command
line or an INCLUDE command is executed. In both cases, when the end of file
is encountered, the parser will again wait for input from stdin. In order
to execute a file and quit in the end of the file, put an EXIT command as
the last command in the file.
$
Overloading Plus
The + operator is overloaded above the following domains:
NumericType + NumericType -> NumericType
VectorType + VectorType -> VectorType (Vector addition)
MatrixType + MatrixType -> MatrixType (Matrix addition)
PolygonType + PolygonType -> PolygonType (Boolean UNION operation)
CurveType + CurveType -> CurveType (Curve curve profiling)
CurveType + CtlPtType -> CurveType (Curve control point profiling)
CtlPtType + CtlPtType -> CurveType (Control points profiling)
ListType + ListType -> ListType (Append lists operator)
StringType + StringType -> StringType (String concat)
StringType + RealType -> StringType (String concat, real as int string)
Note: Boolean UNION of two disjoint objects (no common volume) will result
with the two objects combined. It is the USER responsibility to make sure that
the non intersecting objects are also disjoint - this system only tests for
no intersection. Boolean UNION of two polyline objects will merge the list
of polylines.
$
Overloading Minus
The - operator is overloaded above the following domains:
As a binary operator:
NumericType - NumericType -> NumericType
VectorType - VectorType -> VectorType (Vectoric difference)
MatrixType - MatrixType -> MatrixType (Matrix difference)
PolygonType - PolygonType -> PolygonType (Boolean SUBTRACT operation)
As a unary operator:
- NumericType -> NumericType
- VectorType -> VectorType (Scale vector by -1)
- MatrixType -> MatrixType (Scale matrix by -1)
- PolygonType -> PolygonType (Boolean NEGATION operation)
- CurveType -> CurveType (Curve parameterization is reversed)
- SurfaceType -> SurfaceType (Surface parameterization is reversed)
Note: Boolean SUBTRACT of two disjoint objects (no common volume) will result
with an empty object. For both a curve and a surface parameterization, reverse
operation (binary minus) causes the object normal to be flipped as a side
effect.
$
Overloading Astrisk
The * operator is overloaded above the following domains:
NumericType * NumericType -> NumericType
VectorType * NumericType -> VectorType (Vector scaling)
VectorType * CurveType -> CurveType (Inner product projection)
VectorType * SurfaceType -> SurfaceType (Inner product projection)
VectorType * VectorType -> NumericType (Inner product)
MatrixType * NumericType -> MatrixType (Matrix Scaling)
MatrixType * PointType -> PointType (Point transformation)
MatrixType * CtlPtType -> CtlPtType (Ctl Point transformation)
MatrixType * VectorType -> VectorType (Vector transformation)
MatrixType * MatrixType -> MatrixType (Matrix multiplication)
MatrixType * GeometricType -> GeometricType (Object transformation)
MatrixType * ListType -> ListType (Object hierarchy transform.)
PolygonType * PolygonType -> PolygonType (Boolean INTERSECTION operation)
InstanceType * MatrixType -> InstanceType (Transform of Instance's matrix)
Note: Boolean INTERSECTION of two disjoint objects (no common volume) will
result with an empty object. Object hierarchy transform transforms any
transformable object (GeometricType) found in the list recursively. Boolean
INTERSECTION of two planar (XY plane) polyline objects will compute the
intersection points of the two lists of polylines.
$
Overloading Slash
The / operator is overloaded above the following domains:
NumericType / NumericType -> NumericType
PolygonType / PolygonType -> PolygonType (Boolean CUT operation)
Note: Boolean CUT of two disjoint objects (no common volume) will result
with an empty object.
$
Overloading Caret
The ^ operator is overloaded above the following domains:
NumericType ^ NumericType -> NumericType
VectorType ^ VectorType -> VectorType (Cross product)
MatrixType ^ NumericType -> MatrixType (Matrix to the (int) power)
PolygonType ^ PolygonType -> PolygonType (Boolean MERGE operation)
StringType ^ StringType -> StringType (String concat)
StringType ^ RealType -> StringType (String concat, real as real string)
Note: Boolean MERGE simply merges the two sets of polygons without any
intersection tests. Matrix powers must be positive integers or -1, in which
case the matrix inverse (if it exists) is computed.
$
Overloading Equal (Assignments)
Assignments are allowed as side effects, in any place in an expression.
If "Expr" is an expression, then "var = Expr" is the exact same expression
with the side effect of setting Var to that value. There is no guarantee
on the order of evaluation, so using Vars that are set within the same
expression is a bad practice. Use parentheses to force the order of
evaluation, i.e., "( var = Expr )".
$
Comparison operators - ==, !=, <, >, <=, >=
The conditional comparison operators can be applied to the following
domains (o for a comparison operator):
NumericType o NumericType -> NumericType
StringType o StringType -> NumericType
PointType o PointType -> NumericType
VectorType o VectorType -> NumericType
PlaneType o PlaneType -> NumericType
The returned NumericType is non-zero if the condition holds, or zero if
not.
For PointTypes, VectorTypes, and PlaneTypes, only == and != comparisons
are valid. This is either the same or different.
For NumericTypes and StringTypes (uses strcmp) all comparisons are valid.
$
Logical Operators - &&, ||, |
Complex logical expressions can be defined using the logical and (&&),
logical or (||) and logical not (!). These operators can be applied
to NumericTypes that are considered Boolean results. That is, true for a
non-zero value, and false otherwise.
The returned NumericType is true if both operands are true for the and
operator, at least one is true for the or operator, and the operand is
false for the not operator. In all other cases, a false is returned.
To make sure Logical expressions are readable, the and and or
operators are defined to have the same priority. Use parentheses to
disambiguate a logical expression and to make it more readable.
$
Geometric Boolean Operations
The IRIT solid modeling system supports Boolean operations between
polyhedra objects. Freeform objects will be automaticaly converted to
a polygonal representation when used in Boolean operations.
The +, *, and - are overloaded to denote Boolean union, intersection
and subtraction when operating on geometric entities. - can also be
used as an unary operator to reverse the object orientation inside out.
IRIT supports Boolean operations on polyhedra models. A polyhedra
based model is simply a collection of polygons. While a polyhedra is
simply a set of polygons, this set must conform to certain conditions:
Every polygon has known adjacent polygons, for all its edges.
Every polygon has a normal that points into the model. That
normal is inside/outside consistent with its adjacent polygons.
In other words, for every polygon, one can locally determine the
inside or the outside of the model. Moreover, every polygon has
neighbors for all its edges, forming a closed object that
consistently delineates inside from outside, globally.
The Boolean operations are set operations conducted between two
such models, M_1 and M_2, that delineate inside from
outside. Boolean Union, Boolean Intersection and Boolean Subtraction
are the three common operations that resemble the exact semantic that
is expected, when treating M_1 and M_2 as three
dimensional point sets.
The Boolean operations can be formulated into a binary tree structure
also known as Constructive Solid Geometry (CSG) tree.
Example:
resolution = 20;
B = box(vector(-1, -1, -0.25), 2, 1.2, 0.5);
C = con2(vector(0, 0, -1.5), vector(0, 0, 3), 0.7, 0.3);
D = convex(B - C);
E = convex(C - B);
F = convex(B + C);
G = convex(B * C);
tr = rotx( -90 ) * roty( 40 ) * rotx( -30 );
All = list( D * tr * trans( vector( 0.6, 0.5, 0.0 ) ),
E * tr * trans( vector( 3.0, 0.0, 0.0 ) ),
F * tr * trans( vector( -2.0, 0.0, 0.0 ) ),
G * tr * trans( vector( 0.7, -1.0, 0.0 ) ) )
* scale( vector( 0.25, 0.25, 0.25 ) )
* trans( vector( -0.1, -0.3, 0.0 ) );
view_mat = rotx( 0 );
view( list( view_mat, All ), on );
save( "booleans", list( view_mat, All ) );
A complete example to compute the union, intersection and both
differences of a box and a truncated cone.
Special cases can be very difficult to handle when considering Boolean
operations. Consider an axes parallel bounding cube. Consider a
second cube rotated alpha degrees from the first cube. At large
angles, the Boolean operations are fairly simple to
compute. Nevertheless, as alpha approaches zero, the almost coplanar
planes of the two intersecting cubes make it very difficult to
robustly and consistently compute their intersection.
There are several flags to control the Boolean operations. See
IRITSTATE command for the "InterCrv", "Coplanar", and "PolySort"
states.
$
Priority of operators
The following table lists the priority of the different operators.
Lowest Operator Name of operator
priority , comma
: colon
&&, || logical and, logical or
=,==,!=,<=,>=,<,> assignment, equal, not equal, less
equal, greater equal, less, greater
+, - plus, minus
*, / multiply, divide
Highest ^ power
priority -, ! unary minus, logical not
The grammar of the IRIT parser follows similar guidelines as
the C language for simple expressions. However, complex statements differ.
See the IF, FOR, FUNCTION, and PROCEDURE below for the usage of these
clauses.
$
Function Description
$
NumericType returning functions
$
ABS
NumericType ABS( NumericType Operand )
Returns the absolute value of the given Operand.
$
ACOS
NumericType ACOS( NumericType Operand )
Returns the arc cosine value (in radians) of the given Operand.
$
AREA
NumericType AREA( PolygonType Object )
Returns the area of the given Object (in object units). Returned is
the area of the polygonal object, not the area of the primitive it might
approximate.
This means that the area of a polygonal approximation of a sphere will be
returned, not the exact area of the sphere.
$
ASIN
NumericType ASIN( NumericType Operand )
Returns the arc sine value (in radians) of the given Operand.
$
ATAN
NumericType ATAN( NumericType Operand )
Returns the arc tangent value (in radians) of the given Operand.
$
ATAN2
NumericType ATAN2( NumericType Operand1, NumericType Operand2 )
Returns the arc tangent value (in radians) of the given ratio:
Operand1 / Operand2, over the whole circle.
$
COS
NumericType COS( NumericType Operand )
Returns the cosine value of the given Operand (in radians).
$
CLNTEXEC
NumericType CLNTEXEC( StringType ClientName )
Initiate communication channels to a client named ClientName.
ClientName is executed by this function as a sub process and
two communication channels are opened between the IRIT server and the
new client, for read and write. See also CLNTREAD, CLNTWRITE, and
CLNTCLOSE. if ClientName is an empty string, the user is provided
with the new communication port to be used and the server blocks for the
user to manualy executed the client after setting the proper
IRIT_SERVER_HOST/PORT environment variables.
Example:
h1 = CLNTEXEC( "" );
h2 = CLNTEXEC( "nuldrvs -s-" );
executes two clients, one is named nuldrvs and the other one is
prompted for by the user. As a result of the second invokation of
CLNTEXEC, the user will be prompted with a message similar to,
Irit: Startup your program - I am waiting...
setenv IRIT_SERVER_PORT 2182
and he/she will need to set the proper environment variable and execute
their client manually.
$
CPOLY
NumericType CPOLY( PolygonType Object )
Returns the number of polygons in the given polygonal Object.
$
DSTPTLN
NumericType DSTPTLN( PointType Pt, PointType LineOrig, VectorType LineRay )
Returns the distance between a given point Pt and line
LineOrig, LineRay. See also PTPTLN.
$
DSTPTPLN
NumericType DSTPTPLN( PointType Pt, PlaneType Plane )
Returns the distance between a given point Pt and plane
Plane.
$
DSTLNLN
NumericType DSTLNLN( PointType Line1Orig, VectorType Line1Ray,
PointType Line2Orig, VectorType Line2Ray )
Returns the distance between two lines defined by point LineiOrig
and ray LineiRay. See also PTSLNLN.
$
EXP
NumericType EXP( NumericType Operand )
Returns the natural exponent value of the given Operand.
$
FLOOR
NumericType FLOOR( NumericType Operand )
Returns the largest integer not greater than Operand.
$
FMOD
NumericType FMOD( NumericType Operand, NumericType Mod )
Returns the floating point remainder of the division of Operand by
Mod.
$
LN
NumericType LN( NumericType Operand )
Returns the natural logarithm value of the given Operand.
$
LOG
NumericType LOG( NumericType Operand )
Returns the base 10 logarithm value of the given Operand.
$
MESHSIZE
NumericType MESHSIZE( SurfaceType Srf, ConstantType Direction )
Returns the size of Srf's mesh in Direction, which
is one of COL or ROW.
Example:
RSize = MESHSIZE( Sphere, ROW );
CSize = MESHSIZE( Sphere, COL );
$
POWER
NumericType POWER( NumericType Operand, NumericType Exp )
Returns Operand to the power of Exp.
$
RANDOM
NumericType RANDOM( NumericType Min, NumericType Max )
Returns a randomized value between Min and Max.
$
SIN
NumericType SIN( NumericType Operand )
Returns the sine value of the given Operand (in radians).
$
SIZEOF
NumericType SIZEOF( ListType List | PolygonType Poly |
CurveType Crv | StringType Str )
Returns the length of a list if List, the number of polygons
if Poly, the length of the control polygon if Crv, or the
number of characters in string if Str.
If, however, only one polygon is in Poly, it returns the number
of vertices in that polygon.
Example:
len = SIZEOF( list( 1, 2, 3 ) );
numPolys = SIZEOF( axes );
numCtlpt = SIZEOF( circle( vector( 0, 0, 0 ), 1 ) );
will assign the value of 3 to the variable len, set numPolys
to the number of polylines in the axes object, and set numCtlPt to 9, the
number of control points in a circle.
$
SQRT
NumericType SQRT( NumericType Operand )
Returns the square root value of the given Operand.
$
TAN
NumericType TAN( NumericType Operand )
Returns the tangent value of the given Operand (in radians).
$
THISOBJ
NumericType THISOBJ( StringType Object )
Returns the object type of the given name of an Object. This can
be one of the constants,
UNDEF_TYPE VECTOR_TYPE MATRIX_TYPE SURFACE_TYPE
NUMERIC_TYPE POINT_TYPE POLY_TYPE TRIMSRF_TYPE
STRING_TYPE CTLPT_TYPE CURVE_TYPE TRIVAR_TYPE
This is also a way to ask if an object by a given name do exist (if the
returned type is UNDEF_TYPE or not).
$
VOLUME
NumericType VOLUME( PolygonType Object )
Returns the volume of the given Object (in object units). It returns
the volume of the polygonal object, not the volume of the object it might
approximate.
This routine decomposes all non-convex polygons to convex ones as a side
effect (see CONVEX).
$
GeometricType returning functions
$
ADAPISO
CurveType ADAPISO( SurfaceType Srf, NumericType Dir, NumericType Eps,
NumericType FullIso, NumericType SinglePath )
Constructs a coverage to Srf using isocurve in the Dir direction,
so that for any point p on surface Srf, there exists a point on one of
the isocurves that is close to p within Eps. If FullIso, the
extracted isocurves span the entire surface domain, otherwise they may
span only a subset of the domain. If SinglePath, an approximation to
a single path (Hamiltonian path) that visits all isocurves is constructed
(not supported). See also COVERPT, COVERISO.
srf = sbezier( list( list( ctlpt( E3, -0.5, -1.0, 0.0 ),
ctlpt( E3, 0.4, 0.0, 0.1 ),
ctlpt( E3, -0.5, 1.0, 0.0 ) ),
list( ctlpt( E3, 0.0, -0.7, 0.1 ),
ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 0.0, 0.7, -0.2 ) ),
list( ctlpt( E3, 0.5, -1.0, 0.1 ),
ctlpt( E3, -0.4, 0.0, 0.0 ),
ctlpt( E3, 0.5, 1.0, -0.2 ) ) ) );
aiso = ADAPISO( srf, COL, 0.1, FALSE, FALSE );
Constructs an adaptive isocurve approximation with tolerance of 0.1 to
surface srf in direction COL. Isocurves are allowed to span a
subset of the surface domain. No single path is needed.
The SinglePath option is currently not supported.
$
ARC
CurveType ARC( VectorType StartPos, VectorType Center, VectorType EndPos )
Constructs an arc between the two end points StartPos and EndPos,
centered at Center. Arc will always be less than 180 degrees, so the
shortest circular path from StartPos to EndPos is selected. The
case where StartPos, Center, and EndPos are collinear is
illegal, since it attempts to define a 180 degrees arc. Arc is constructed
as a single rational quadratic Bezier curve.
Example:
Arc1 = ARC( vector( 1.0, 0.0, 0.0 ),
vector( 1.0, 1.0, 0.0 ),
vector( 0.0, 1.0, 0.0 ) );
constructs a 90 degrees arc, tangent to both the X and Y axes at coordinate 1.
$
AOFFSET
CurveType AOFFSET( CurveType Crv, NumericType OffsetDistance,
NumericType Epsilon, NumericType TrimLoops,
NumericType BezInterp )
Computes an offset of OffsetDistance with globally bounded error
(controlled by Epsilon). The smaller Epsilon is, the better
the approximation to the offset. The bounded error is achieved by adaptive
refinement of the Crv.
If TrimLoops is TRUE or on, the regions of the object that
self-intersect as a result of the offset operation are trimmed away.
If BezInterp is TRUE, each curve's segment is interpolated instead
of approximated.
Example:
OffCrv1 = AOFFSET( Crv, 0.5, 0.01, FALSE, FALSE );
OffCrv2 = AOFFSET( Crv, 0.5, 0.01, TRUE, FALSE );
computes an adaptive offset to Crv with OffsetDistance of
0.5 and Epsilon of 0.01 and trims the self-intersection loops in
the second instrance.
See also OFFSET, LOFFSET, and MOFFSET.
$
BBOX
ListType BBOX( GeometricTreeType Geom )
Given a (tree of ) geometry, Geom computes its bounding box and
return it as a list of six numbers: XMin/Max, YMin/Max, ZMin/Max, in this
order.
Example:
B1 = BBOX( axes );
$
BOOLONE
SurfaceType BOOLONE( CurveType Crv )
Given a closed curve, the curve is subdivided into four segments equally
spaced in the parametric space that are fed into BOOLSUM. Useful if
a surface should "fill" the area enclosed by a closed curve.
Example:
Srf = BOOLONE( circle( vector( 0.0, 0.0, 0.0 ), 1.0 ) );
Creates a disk surface containing the area enclosed by the unit circle.
$
BOOLSUM
SurfaceType BOOLSUM( CurveType Crv1, CurveType Crv2,
CurveType Crv3, CurveType Crv4 )
Construct a surface using the provided four curves as its four boundary
curves. Curves do not have to have the same order or type, and will be
promoted to their least common denominator. The end points of the four
curves should match as follows:
Crv1 start point, to Crv3 start point.
Crv1 end point, to Crv4 start point.
Crv2 start point, to Crv3 end point.
Crv2 end point, to Crv4 end point.
where Crv1 and Crv2 are the two boundaries in one parametric
direction, and Crv3 and Crv4 are the two boundaries in the other
parametric direction.
Example:
Cbzr1 = cbezier( list( ctlpt( E3, 0.1, 0.1, 0.1 ),
ctlpt( E3, 0.0, 0.5, 1.0 ),
ctlpt( E3, 0.4, 1.0, 0.4 ) ) );
Cbzr2 = cbezier( list( ctlpt( E3, 1.0, 0.2, 0.2 ),
ctlpt( E3, 1.0, 0.5, -1.0 ),
ctlpt( E3, 1.0, 1.0, 0.3 ) ) );
Cbsp3 = cbspline( 4,
list( ctlpt( E3, 0.1, 0.1, 0.1 ),
ctlpt( E3, 0.25, 0.0, -1.0 ),
ctlpt( E3, 0.5, 0.0, 2.0 ),
ctlpt( E3, 0.75, 0.0, -1.0 ),
ctlpt( E3, 1.0, 0.2, 0.2 ) ),
list( KV_OPEN ) );
Cbsp4 = cbspline( 4,
list( ctlpt( E3, 0.4, 1.0, 0.4 ),
ctlpt( E3, 0.25, 1.0, 1.0 ),
ctlpt( E3, 0.5, 1.0, -2.0 ),
ctlpt( E3, 0.75, 1.0, 1.0 ),
ctlpt( E3, 1.0, 1.0, 0.3 ) ),
list( KV_OPEN ) );
Srf = BOOLSUM( Cbzr1, Cbzr2, Cbsp3, Cbsp4 );
$
BOX
PolygonType BOX( VectorType Point,
NumericType Dx, NumericType Dy, NumericType Dz )
Creates a BOX polygonal object, whose boundary is coplanar with the
XY, XZ, and YZ planes. The BOX is defined by Point as
base position, and Dx, Dy, Dz as BOX dimensions. Negative dimensions
are allowed.
Example:
B = BOX( vector( 0, 0, 0 ), 1, 1, 1);
creates a unit cube from 0 to 1 in all axes.
$
BZR2BSP
CurveType BZR2BSP( CurveType Crv )
or
SurfaceType BZR2BSP( SurfaceType Srf )
Creates a Bspline curve or a Bspline surface from the given Bezier curve or
Bezier surface. The Bspline curve or surface is assigned open end knot
vector(s) with no interior knots, in the parametric domain of zero to one.
Example:
BspSrf = BZR2BSP( BzrSrf );
$
BSP2BZR
CurveType | ListType BSP2BZR( CurveType Crv )
or
SurfaceType | ListType BSP2BZR( SurfaceType Srf )
Creates Bezier curve(s) or surface(s) from a given Bspline curve or a
Bspline surface. The Bspline input is subdivided at all internal knots to
create Bezier curves or surfaces. Therefore, if the input Bspline does have
internal knots, a list of Bezier curves or surfaces is returned. Otherwise,
a single Bezier curve or surface is returned.
Example:
BzrCirc = BSP2BZR( circle( vector( 0.0, 0.0, 0.0 ), 1.0 ) );
would subdivide the unit circle into four 90 degrees Bezier arcs returned
in a list.
$
CBEZIER
CurveType CBEZIER( ListType CtlPtList )
Creates a Bezier curve out of the provided control point list. CtlPtList
is a list of control points, all of which must be of type (E1-E5, P1-P5),
or regular PointType defining the curve's control polygon. Curve's point type
will be of a space which is the union of the spaces of all points.
Example:
s45 = sin(pi / 4);
Arc90 = CBEZIER( list( ctlpt( P2, 1.0, 0.0, 1.0 ),
ctlpt( P2, s45, s45, s45 ),
ctlpt( P1, 1.0, 1.0 ) ) );
constructs an arc of 90 degrees as a rational quadratic Bezier curve.
$
CBISECTOR
{ CurveType | SurfaceType } CBISECTOR( CurveType Crv,
NumericType UseNrmlTan,
NumericType Tolerance,
NumericType NumerImprove,
NumericType SameNormal )
or
{ CurveType | SurfaceType } CBISECTOR( ListType TwoCrvs,
NumericType UseNrmlTan,
NumericType Tolerance,
NumericType NumerImprove,
NumericType SameNormal )
or
SurfaceType CBISECTOR( ListType CrvPt,
NumericType UseNrmlTan,
NumericType Tolerance,
NumericType NumerImprove,
NumericType SameNormal )
Computes the self bisector curve(s) for Crv or the bisector(s)
of TwoCrvs or the bisector of a curve and a point, CrvPt. If
UseNrmlTan is equal to zero or one, the bisector curves are computed
using angular equality of tangents or normals of the given curve or curves,
respectively. Tolerance controls the accuracy of the computation,
with 10 as a good starting value. If Tolerance is negative,
NumerImprove can be either TRUE or FALSE, allowing or disabling a
final numerical imrpovement stage. SameNormal can also assume a
TRUE or FALSE value, selecting only opposite facing normals, if TRUE.
Nevertheless, if UseNrmlTan equals -1, the bisector surface of
the given two curves TwoCrvs or a curve and a point CrvPt is
computed analytically. For two 3-space curves this is indeed the bisector
surface while for two 2-space curves, the zero set of the returned surface
provides the bisector curve. If UseNrmlTan equals -1, then
Tolerance, NumerImprove, and SameNormal are all ignored.
If UseNrmlTan equals -2 and the two curves are E2, a different
bisector surface is computed whose zero set provides the bisector curve.
If UseNrmlTan equals -2, then Tolerance, NumerImprove,
and SameNormal are all ignored.
If UseNrmlTan equals -3 and the two curves are 3-space curves,
an alpha bisector surface is computed, whose ratio is set via Tolerance
relative to NumerImprove, such that the sum of Tolerance +
NumerImprove is not zero. If UseNrmlTan equals -3, then
SameNormal is ignored. See also SBISECTOR.
Example:
c1 = cbezier( list( ctlpt( E2, -0.5, -0.2 ),
ctlpt( E2, 0.0, -0.2 ),
ctlpt( E2, 0.6, 0.6 ) ) );
c2 = cbezier( list( ctlpt( E2, 0.3, -0.7 ),
ctlpt( E2, -0.2, -0.7 ),
ctlpt( E2, 0.7, 0.6 ) ) );
BisectCrvs = CBISECTOR( list( c1, c2 ), 0, 12, true, false );
All = list( c1, c2, BisectCrvs );
interact( list( All, view_mat2d ) );
c1 = creparam( pcircle( vector( 0.0, 0.0, 0.0 ), 1 ), 0, 1 );
c2 = cbezier( list( ctlpt( E3, -1.0, 0.0, 1.0 ),
ctlpt( E3, 1.0, 0.0, -1.0 ) ) );
BisectSrf = CBISECTOR( list( c1, c2 ), -1, -1, true, false );
interact( list( c1, c2, BisectSrf ) );
computes two bisectors, one for to planar curves as a set of bisector curves,
and the other as a bisector surfce of a Z parallel line and a circle in the
XY plane.
$
CBSPLINE
CurveType CBSPLINE( NumericType Order, ListType CtlPtList,
ListType KnotVector )
Creates a Bspline curve out of the provided control point list, the
knot vector, and the specified order. CtlPtList is a list of control
points, all of which must be of type (E1-E5, P1-P5), or regular PointType
defining the curve's control polygon. Curve's point type will be of a space
which is the union of the spaces of all points.
The length of the KnotVector must be equal to the number of control
points in CtlPtList plus the Order. If, however, the length of
the knot vector is equal to #CtlPtList + Order + Order - 1 the curve
is assumed periodic.
The knot vector list may be specified as either list( KV_OPEN ) or
list( KV_FLOAT ) or list( KV_PERIODIC ) in which a uniform
open, uniform floating or uniform periodic knot vector with the appropriate
length is automatically constructed.
Example:
s45 = sin(pi / 4);
HalfCirc = CBSPLINE( 3,
list( ctlpt( P3, 1.0, 1.0, 0.0, 0.0 ),
ctlpt( P3, s45, s45, s45, 0.0 ),
ctlpt( P3, 1.0, 0.0, 1.0, 0.0 ),
ctlpt( P3, s45, -s45, s45, 0.0 ),
ctlpt( P3, 1.0, -1.0, 0.0, 0.0 ) ),
list( 0, 0, 0, 1, 1, 2, 2, 2 ) );
constructs an arc of 180 degrees in the XZ plane as a rational quadratic
Bspline curve.
Example:
c = CBSPLINE( 4,
list( ctlpt( E2, 0.5, 0.5 ),
ctlpt( E2, -0.5, 0.5 ),
ctlpt( E2, -0.5, -0.5 ),
ctlpt( E2, 0.5, -0.5 ) ),
list( KV_PERIODIC ) );
color( c, red );
viewobj( c );
c1 = cregion( c, 3, 4 );
color( c1, green );
c2 = cregion( c, 4, 5 );
color( c2, yellow );
c3 = cregion( c, 5, 6 );
color( c3, cyan );
c4 = cregion( c, 6, 7 );
color( c3, magenta );
viewobj( list( c1, c2, c3, c4 ) );
creates a periodic curve and extracts its four polynomial domains as four
open end Bspline curves.
$
CCINTER
ListType CCINTER( CurveType Crv1, CurveType Crv2, NumericType Epsilon,
NumericType SelfInter )
or
SurfaceType CCINTER( CurveType Crv1, CurveType Crv2, NumericType Epsilon,
NumericType SelfInter )
Computes the intersection point(s) of Crv1 and Crv2 in the
XY plane. Since this computation involves numeric operations, Epsilon
controls the accuracy of the parametric values of the result.
It returns a list of PointTypes, each containing the parameter of Crv1
in the X coordinate, and the parameter of Crv2 in the Y coordinate.
If, however, Epsilon is negative, a scalar field surface representing
the square of the distance function is returned instead.
If SelfInter is TRUE, Crv1 and Crv2 can be the same
curve, and self-intersection points are searched instead.
Example:
crv1 = cbspline( 3,
list( ctlpt( E2, 0, 0 ),
ctlpt( E2, 0, 0.5 ),
ctlpt( E2, 0.5, 0.7 ),
ctlpt( E2, 1, 1 ) ),
list( KV_OPEN ) );
crv2 = cbspline( 3,
list( ctlpt( E2, 1, 0 ),
ctlpt( E2, 0.7, 0.25 ),
ctlpt( E2, 0.3, 0.5 ),
ctlpt( E2, 0, 1 ) ),
list( KV_OPEN ) );
inter_pts = CCINTER( crv1, crv2, 0.0001, FALSE );
Computes the parameter values of the intersection point of crv1 and
crv2 to a tolerance of 0.0001.
$
CCRVTR
NumericType CCRVTR( CurveType Crv, NumericType Epsilon )
or
CurveType CCRVTR( CurveType Crv, NumericType Epsilon )
Computes the extreme curvature points on Crv in the XY plane.
This set includes not only points of maximum (convexity) and mimumum
(concavity) curvature, but also points of zero curvature such as
inflection points.
Since this operation is partially numeric, Epsilon is used to set
the needed accuracy. It returns the parameter value(s) of the location(s)
with extreme curvature along the Crv.
If, however, Epsilon is negative, the curvature scalar field
curve is returned as a two dimensional rational vector field curve, for
which the first dimension is equal to the parameter, and the second is the
curvature value at that parameter.
This function computes the curvature scalar field for planar curves as,
x' y'' - x'' y'
k(t) = ----------------
2 2 3/2
( x' + y' )
and computes kN for three dimensional curves as the following vector field,
C' x C'' C' (C' x C'') x C'
k(t) N(t) = K(t) B(t) x T(t) = -------- x ----- = ---------------
3 | C' | 4
| C'| | C' |
The extremum values are extracted from the computed curvature field.
This curvature field is a high order curve, even if the input geometry is
of low order. This is especially true for rational curves, for which the
quotient rule for differentiation is used and almost doubles the degree
in every differentiation.
See also CZEROS, CEXTREMES, and SCRVTR.
Example:
crv = cbezier( list( ctlpt( E2, -1.0, 0.5 ),
ctlpt( E2, -0.5, -2.0 ),
ctlpt( E2, 0.0, 1.0 ),
ctlpt( E2, 1.0, 0.0 ) ) ) * rotz( 30 );
crvtr = CCRVTR( crv, 0.001 );
pt_crvtr = nil();
pt = nil();
for ( i = 1, 1, sizeof( crvtr ),
( pt = ceval( crv, nth( crvtr, i ) ) ):
snoc( pt, pt_crvtr )
);
interact( list( crv, pt_crvtr ) );
finds the extreme curvature points in Crv and displays them all
with the curve.
$
CDERIVE
CurveType CDERIVE( CurveType Curve )
Returns a vector field curve representing the differentiated curve,
also known as the Hodograph curve.
Example:
Circ = circle( vector( 0.0, 0.0, 0.0 ), 1.0 );
Hodograph = CDERIVE( Circ );
$
CDIVIDE
ListType CDIVIDE( CurveType Curve, NumericType Param )
Subdivides a curve into two sub-curves at the specified parameter value.
Curve can be either a Bspline curve in which Param must be
within the Curve's parametric domain, or a Bezier curve in which Param
can be arbitrary, extrapolating if not in the range of zero to one.
Returns a list of the two sub-curves. The individual curves may be
extracted from the list using the NTH command.
Example:
CrvLst = CDIVIDE( Crv, 1.3 );
Crv1 = nth( CrvLst, 1 );
Crv2 = nth( CrvLst, 2 );
subdivides the curve Crv at the parameter value of 0.5.
$
CEDITPT
CurveType CEDITPT( CurveType Curve, CtlPtType CtlPt, NumericType Index )
Provides a simple mechanism to manually modify a single control point number
Index (base count is 0) in Curve, by substituting CtlPt
instead. CtlPt must have the same point type as the control points of
the Curve. Original curve Curve is not modified.
Example:
CPt = ctlpt( E3, 1, 2, 3 );
NewCrv = CEDITPT( Curve, CPt, 1 );
constructs a NewCrv with the second control point of Curve being
CPt.
$
CENVOFF
SurfaceType CENVOFF( CurveType Curve,
NumericType Height, NumericType Tolerance )
or
ListType CENVOFF( CurveType Curve, NumericType Height, NumericType Tolerance )
Computes the offset envelope of a given planar curve Curve. The
offset envelope is the envelope of cones with apex on point on Curve
in the Z direction. Height is the height of the cone which also
equals the offset distance or the width of the cones. Tolerance
controls the accuracy of the offset approximation.
If Curve is closed, two surfaces are created in the offset envelope,
one for the inside and one for the outside. If Curve is open, a single
envelope offset surface is computed wrapping around both sides.
Example:
c1 = cbezier( list( ctlpt( E2, -0.8, 0.0 ),
ctlpt( E2, -0.2, 1.0 ),
ctlpt( E2, 0.2, 0.0 ),
ctlpt( E2, 0.8, 0.6 ) ) );
s1 = CENVOFF( c1, 0.5, 0.01 );
Computes an envelope offset surface for a cubic Bezier curve c1 of
Height of 0.5 and Tolerance of 0.01.
$
CEVAL
CtlPtType CEVAL( CurveType Curve, NumericType Param )
Evaluates the provided Curve at the given Param value.
Param should be in the curve's parametric domain if Curve is
a Bspline curve, or between zero and one if Curve is a Bezier curve.
The returned control point has the same point type as the control points
of the Curve.
Example:
CPt = CEVAL( Crv, 0.25 );
evaluates Crv at the parameter value of 0.25.
$
CEXTREMES
ListType CEXTREMES( CurveType Crv, NumericType Epsilon, NumericType Axis )
Computes the extreme set of the given Crv in the given axis (1 for X,
2 for Y, 3 for Z). Since this computation is numeric, an Epsilon is
also required to specify the desired tolerance. It returns a list of
all the parameter values (NumericType) in which the curve takes an extreme
value.
Example:
extremes = CEXTREMES( Crv, 0.0001, 1 );
Computes the extreme set of curve crv, in the X axis, with
error tolerance of 0.0001. See also CZERO.
$
CINFLECT
ListType CINFLECT( CurveType Crv, NumericType Epsilon )
or
CurveType CINFLECT( CurveType Crv, NumericType Epsilon )
Computes the inflection points of Crv in the XY plane.
Since this computation is numeric, an Epsilon is also required
to specify the desired tolerance. It returns a list of all the
parameter values (NumericType) in which the curve has an inflection point.
If, however, Epsilon is negative, a scalar field curve representing
the sign of the curvature of the curve is returned instead.
The sign of curvature scalar field is equal to
s(t) = x' y'' - x'' y'
Example:
inflect = CINFLECT( crv, 0.001 );
pt_inflect = nil();
pt = nil();
for ( i = 1, 1, sizeof( inflect ),
pt = ceval( crv, nth( inflect, i ) ):
snoc( pt, pt_inflect )
);
interact( list( axes, crv, pt_inflect ), 0);
Computes the set of inflection points of curve crv with error
tolerance of 0.001. This set is then scanned in a loop and
evaluated to the curve's locations which are then displayed with the crv.
See also CZEROS, CEXTREMES, and CCRVTR.
$
CINTERP
CurveType CINTERP( ListType PtList, NumericType Order, NumericType Size,
ConstantType Param, NumericType Periodic)
Computes a Bspline polynomial curve that interpolates or approximates
the list of points in PtList. The Bspline curve will have order
Order and Size control points, and will be periodic if
periodic is none zero. The knots will be spaced according to
Param which can be one of PARAM_UNIFORM, PARAM_CHORD or
PARAM_CENTRIP. The former prescribes a uniform knot sequence and the
latters specify knot spacing according to the chord
length and a square root of the chord length. A periodic curve will be
coerced to have PARAM_UNIFORM knot sequence.
Use of Periodic end conditions can create cases with degenerated
linear systems (determinant equal zero). Increase or decrease of the
Order of the Bspline by one will resolve the problem.
All points in PtList must be of type (E1-E5, P1-P5) control point,
or regular PointType. If Size is equal to the number of points in
PtList the resulting curve will interpolate the data set.
Otherwise, if Size is less than the number of points in PtList
the point data set will be least square approximated. In no time can Size
be lower than Order. Size of zero forces interpolation by
setting Size to the data set size.
All interior knots will be distinct preserving maximal continuity.
The resulting Bspline curve will have open end conditions.
Example:
pl = nil();
for ( x = 0, 1, 100,
snoc(point(cos(x / 5), sin(x / 5), x / 50 - 1), pl)
);
c = CINTERP( pl, 3, 21, PARAM_UNIFORM );
Samples a helical curve at 100 points and least square fit a quadratic
Bspline curve with 21 point to the data set. The curve will have a uniform
knot spacing.
$
CIRCLE
CurveType CIRCLE( VectorType Center, NumericType Radius )
Constructs a circle at the specified Center with the specified
Radius. The returned circle is a Bspline curve of four piecewise Bezier
90 degree arcs. The construced circle is always parallel to the XY plane.
Use the linear transformation routines to place the circle in the appropriate
orientation and location.
$
CIRCPOLY
PolygonType CIRCPOLY( VectorType Normal, VectorType Trans, NumericType Radius )
Defines a circular polygon in a plane perpendicular to Normal that
contains the Trans point. Constructed polygon is centered at
Trans. RESOLUTION vertices will be defined with Radius from
distance from Trans.
Alternative ways to construct a polygon are manual construction of the
vertices using POLY, or the construction of a flat ruled surface using
RULEDSRF.
$
CLNTREAD
AnyType CLNTREAD( NumericType Handler, NumericType Block )
Reads one object from a communication channel of a client.
Handler contains the index of the communication channel opened via
CLNTEXEC. If no data is available in the communication channel, this
function will block for at most Block millisecond until data is found
or timeout occurs. In the latter, a single StringType object is returned
with the content of "no data (timeout)". See also CLNTWRITE, CLNTCLOSE,
and CLNTEXEC.
Example:
h2 = clntexec( "xmtdrvs -s-" );
.
.
Model = CLNTREAD( h2 );
.
.
clntclose( h2,TRUE );
reads one object from client through communication channel h2 and save it
in variable Model.
$
CMESH
CurveType CMESH( SurfaceType Srf, ConstantType Direction, NumericType Index )
Returns a single ROW or COLumn as specified by the Direction and
Index (base count is 0) of the control mesh of surface Srf.
The returned curve will have the same knot vector as Srf in the
appropriate direction. See also CSURFACE.
This curve is not necessarily in the surface Srf.
Example:
Crv = CMESH( Srf, COL, 0 );
extracts the first column of surface Srf as a curve. See also
CSURFACE.
$
CMORPH
CurveType CMORPH( CurveType Crv1, CurveType Crv2,
NumericType Method, NumericType Blend )
or
ListType CMORPH( CurveType Crv1, CurveType Crv2,
NumericType Method, NumericType Blend )
Creates a new curve which is a metamorph of the two given curves.
The two given curves must be compatible (see FFCOMPAT) before this blend
is invoked. Very useful if a sequence that "morphs" one curve to another
is to be created. Several methods of metamorphosis are supported according
to the value of Method,
0 Simple convex blend.
1 Corner/Edge cutting scheme, scaled to same curve length.
2 Corner/Edge cutting scheme, scaled to same bounding box.
3 Same as 1 but with filtering out of tangencies.
4 Same as 2 but with filtering out of tangencies.
5 Multiresolution decompsition based metamorphosis. See CMULTRES.
In Method 1, Blend is a number between zero (Crv1)
and one (Crv2) defining the similarity to Crv1 and Crv2,
respectively. A single curve is returned.
In Methods 2 to 5, Blend is a step size for the
metamorphosis operation and a whole list describing the entire
metamorphosis operation is returned.
Examples:
for ( i = 0, 1, 300,
c = CMORPH( crv1a, crv1b, 0, i / 300.0 ):
color( c, yellow ):
viewobj( c )
);
crvs = CMORPH( crv1a, crv1b, 2, 0.003 );
snoc( crv1b, crvs );
for ( i = 1, 1, sizeof( crvs ),
c = nth( crvs, i ):
color( c, yellow ):
viewobj( c )
);
Turtle2 = ffmatch( Wolf, Turtle, 20, 100, 2, false, 2 );
ffcompat( Wolf, Turtle2 );
for ( i = 0, 1, 25,
c = CMORPH( Wolf, Turtle2, 0, i / 25 ):
color( c, yellow ):
viewobj( c )
);
creates three metamorphosis animation sequences, one that is based on
a convex blend, one that is based on corner/edge cutting scheme.
See alost SMORPH, and a third that employs FFMATCH.
$
CMULTIRES
ListType CMULTIRES( CurveType Crv, NumericType Discont )
Computes a multiresolution decomposition of curve Crv using least
squares approximation. The resulting list of curves describes an hierarchy
of curves in linear subspaces of the space Crv was in that can be
sum algebraically to form Crv. Each of the curves in the hierarchy
is a least squares approximation of Crv in the subspace it is defined
in. Discont is a boolean flat that controls the way tangent
discontinuities are computed throughout the decomposition.
Example:
MRCrv = CMULTIRES( Animal, false );
sum = nth( MRCrv, 1 );
MRCrvs = list( sum * tx( 3.0 ) );
for ( ( i = 2 ), 1, sizeof( MRCrv ),
sum = symbsum( sum, nth( MRCrv, i ) ):
snoc( sum * tx( ( 3 - i ) * 1.5 ), MRCrvs )
);
All = MRCrvs * sc ( 0.25 );
view( All, on );
Computes a multiresolution decomposition to curve CrossSec as
MRCrv and display all the decomposition levels by summing them all
up. The use of none as on object name allows one to display an
object in the display device without replacing the previous object in the
display device, carrying the same name.
creates two metamorphosis animation sequences, one that is based on
a convex blend and one that is based on corner/edge cutting scheme.
$
CNORMAL
VectorType CNORMAL( CurveType Crv, NumericType TParam )
Computes the normal vector to curve Crv at the
parameter values TParam. The returned vector has a
unit length.
Example:
Normal = CNORMAL( Crv, 0.5 );
computes the normal to Crv at the parameter value 0.5.
See also CNRMLCRV.
$
CNRMLCRV
CurveType CNRMLCRV( CurveType Crv )
Symbolically computes a vector field curve representing the non-normalized
normals of the given curve. That is a normal vector filed, evaluated at
t, provides a vector in the direction of the normal of the original
curve at t. The normal curve computed is in fact equal to kN where k
is the curvature of Crv and N is its normal.
Example:
NrmlCrv = CNRMLCRV( Crv );
$
CNVXHULL
PolygonType CNVXHULL( PolygonType Poly | PolylineType Poly );
or
CurveType CNVXHULL( CurveType Crv );
Computes the convex hull of the given set of Poly or Curve.
For curves, the result might be partial if the curve is not closed or
periodic. See also CRV2TANS and CRVPTTAN.
Example:
Pts1 = nil();
Pts2 = nil();
for ( i = 0, 1, 7,
R = 0.2 + fmod( i, 2 ) / 2:
Pt = ctlpt( E2, R * cos( i * 2 * pi / 8 ), R * sin( i * 2 * pi / 8 ) ):
snoc( Pt, Pts1 ):
snoc( coerce( Pt, point_type ), Pts2 ) );
Crv = coerce( cbspline( 4, Pts1, list( KV_PERIODIC ) ), KV_OPEN );
CHPts = CNVXHULL( poly( Pts2, 0 ), 0 );
CHCrv = CNVXHULL( Crv, 0 );
computes the convex hull of a given control polygon and freeform curve.
$
COERCE
AnyType COERCE( AnyType Object, ConstantType NewType )
Provides a coercion mechanism between different objects or object types.
PointType, VectorType, PlaneType, CtlPtType can be all coerced to each
other by using the NewType of POINT_TYPE, VECTOR_TYPE, PLANE_TYPE,
or one of E1-E5, P1-P5 (CtlPtType). Similarly, CurveType, SurfaceType,
TrimSrfType, TriSrfType, and TrivarType can all be coerced to hold different
CtlPtType of control points, or even different open end conditions from
KV_PERIODIC to KV_FLOAT to KV_OPEN.
If a scalar (E1 or P1) curve is coerced to E2 or P2 curve or
a scalar (E1 or P1) surface is coerced to E3 or P3 surface, the Y (YZ)
coordinate(s) is (are) updated to hold the parametric domain of the curve
(surface). That is X = U (Y = V).
Example:
CrvE2 = COERCE( Crv, E2 );
coerce Crv to a new curve that will have an E2 CtlPtType control
points. Coerction of a projective curve (P1-P5) to a Euclidean curve
(E1-E5) does not preseve the shape of the curve.
$
COMPOSE
CurveType COMPOSE( CurveType Crv1, CurveType Crv2 )
or
CurveType COMPOSE( SurfaceType Srf, CurveType Crv )
Symbolically compute the composition curve Crv1(Crv2(t)) or
Srf(Crv(t)). In Crv1(Crv2(t), Crv1 can be any curve
while Crv2 must be a one-dimensional curve that is either E1 or
P1. In Srf(Crv(t)), Srf can be any surface, while Crv
must be a two-dimensional curve, that is either E2 or P2. Both Crv2
in the curve's composition, and Crv is the surface's composition
must be contained in the curve or surface parametric domain.
Example:
srf = sbezier( list( list( ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 0.0, 0.5, 1.0 ),
ctlpt( E3, 0.0, 1.0, 0.0 ) ),
list( ctlpt( E3, 0.5, 0.0, 1.0 ),
ctlpt( E3, 0.5, 0.5, 0.0 ),
ctlpt( E3, 0.5, 1.0, 1.0 ) ),
list( ctlpt( E3, 1.0, 0.0, 1.0 ),
ctlpt( E3, 1.0, 0.5, 0.0 ),
ctlpt( E3, 1.0, 1.0, 1.0 ) ) ) );
crv = coerce( circle( vector( 0.0, 0.0, 1.0 ), 0.4 ), p2 ) *
trans( vector( 0.5, 0.5, 0.0 ) );
comp_crv = COMPOSE( srf, crv );
composes a circle Crv to be on the surface Srf.
$
CON2
PolygonType CON2( VectorType Center, VectorType Direction,
NumericType Radius1, NumericType Radius2 )
Creates a truncated CONE geometric object, defined by Center as the
center of the main base of the CONE, Direction as both the CONE's axis
and the length of CONE, and the two radii Radius1/2 of the two bases of
the CONE.
Unlike the regular cone (CONE) constructor which has inherited
discontinuities in its generated normals at the apex, CON2 can be used to
form a (truncated) cone with continuous normals.
See RESOLUTION for the accuracy of the CON2 approximation as a polygonal
model. See also CONE. See IRITSTATE's "PrimRatSrfs" and "PrimRatSrfs"
state variables.
Example:
Cone2 = CON2( vector( 0, 0, -1 ), vector( 0, 0, 4 ), 2, 1 );
constructs a truncated cone with bases parallel to the XY plane at
Z = -1 and Z = 3, and with radii of 2 and 1 respectively.
$
CONE
PolygonType CONE( VectorType Center, VectorType Direction,
NumericType Radius )
Creates a CONE geometric object, defined by Center as the center of
the base of the CONE, Direction as the CONE's axis and height, and
Radius as the radius of the base of the CONE.
See RESOLUTION for accuracy of the CONE approximation as a polygonal model.
Example:
Cone1 = CONE( vector( 0, 0, 0 ), vector( 1, 1, 1 ), 1 );
constructs a cone based in an XY parallel plane, centered at the origin
with radius 1 and with tilted apex at ( 1, 1, 1 ).
See IRITSTATE's "PrimRatSrfs" and "PrimRatSrfs" state variables.
See also CON2.
$
CONTOUR
PolygonType CONTOUR( SurfaceType ContouredSrf, PlaneType ContourPlane )
or
PolygonType CONTOUR( SurfaceType ContouredSrf, PlaneType ContourPlane,
SurfaceType MappedSrf )
Contours surface ContouredSrf by intersecting plane ContourPlane
with a polygonal approximating of ContouredSrf with resolution set
via variable RESOLUTION (10 is a good starting resolutionvalue, 20 is
extreme).
If ContouredSrf is a scalar field surface of type E1 or P1 and
MappedSrf is provided, ContouredSrf is contoured above its
parametric domain (U is X, V is Y) and the resulting parametric curve
is composed with MappedSrf to yield the returned value.
Example:
resolution = 20;
nglass = snrmlsrf( glass ) * vector( 1, 1, 1 );
sils = contour( nglass, plane( 1, 0, 0, 0 ), glass );
color( sils, cyan );
attrib( sils, "dwidth", 4 );
view( list( axes, glass, sils ), on );
Computes the normal field of the surface glass, project it onto
the viewing direction of (1, 1, 1) and contour the resulting scalar field
with the plane X = 0, to extract the silhouette curves from viewing direction
(1, 1, 1).
$
CONVEX
PolygonType CONVEX( PolygonType Object )
or
ListType CONVEX( ListType Object )
Converts non-convex polygons in Object, into convex ones. New vertices
are introduced into the polygonal data during this process. The Boolean
operations require the input to have convex polygons only (although it may
return non convex polygons...) and it automatically converts non-convex input
polygons to convex ones, using this same routine.
However, some external tools (like irit2ray and poly3d-h) require convex
polygons. This function must be used on the objects to guarantee that
only convex polygons are saved into data files for these external tools.
Example:
CnvxObj = CONVEX( Obj );
save( "data", CnvxObj );
converts non-convex polygons into convex ones, so that the data file can be
used by external tools requiring convex polygons.
$
COORD
AnyType COORD( AnyType Object, NumericType Index )
Extracts an element from a given Object, at index Index. From
a PointType, VectorType, PlaneType, CtlPtType and MatrixType, a NumericType
is returned with Index 0 for the X axis, 1 for the Y axis etc.
Index 0 denotes the weight of CtlPtType. For a PolygonType that
contains more than one polygon, the Indexth polygon is returned. For
a PolygonType that contains a single Polygon, the Indexth vertex is
returned. For a CurveType or a SurfaceType, the Indexth CtlPtType is
returned. For a ListType, COORD behaves like NTH and returns the Indexth
object in the list. For a StringType, the Indexth character is returned
as its ASCII numeric code.
Example:
a = vector( 1, 2, 3 );
vector( COORD( a, 0 ), COORD( a, 1 ), COORD( a, 2 ) );
a = ctlpt( P2, 6, 7, 8, 9 );
ctlpt( P3, coord( a, 0 ), coord( a, 1 ), coord( a, 2 ), coord( a, 3 ) );
a = plane( 10, 11, 12, 13 );
plane( COORD( a, 0 ), COORD( a, 1 ), COORD( a, 2 ), COORD( a, 3 ) );
constructs a vector/ctlpt/plane and reconstructs it by extracting the
constructed scalar components of the objects using COORD.
See also COERCE.
$
COVERISO
CurveType COVERISO( TrivarType TV,
NumericType NewOfStrokes,
NumericType StrokeType,
PointType MinMaxPwrLen,
NumericType StepSize,
NumericType IsoVal )
Computes a coverage for an iso surface of a trivariate function TV,
using curves. NewOfStrokes strokes are distributed on the iso surface
with length that is set via MinMaxPwrLen. MinMaxPwrLen is
a triplet of the form (Min, Max, Power) that determines the length of the
strokes as,
Avg = (Max + Min) / 2, Dev = (Max - Min) / 2
Length = Avg + Dev * Random(0, 1) ^ Pwr.
StepSize controls the steps size of the piecewise linear approximation
formed and should be typically smaller than Min.
StrokeType can be one of,
1 Draw strokes along minimal principal curvature.
2 Draw strokes along maximal principal curvature.
3 Draw strokes along both principal curvatures.
4 Draw strokes along constant X planes.
5 Draw strokes along constant Y planes.
6 Draw strokes along constant Z planes.
Finally, IsoVal controls the constant value of the iso surface level.
See also ADAPISO, COVERPT, MRCHCUBE, TVLOAD.
Example:
IsoVal = 0.12;
Cover = CoverIso( ThreeCyls, 500, 1, vector( 3, 10, 1.0 ), 0.2, IsoVal );
draws 500 strokes on the iso surface of trivariate ThreeCyls at iso
value IsoVal and step size of 0.2. Strokes are drawn in length of
3 to 10 along line of curvatures of minimal curvature.
$
COVERPT
PolygonType COVERPT( PolygonType Model,
NumericType NumOfPts,
VectorType ViewDir )
Computes a uniform point distribution on the given polygonal Model.
Approximately NumOfPts points are uniformly distributed on the
Model's surface, provided ViewDir is the zero vector.
If ViewDir is a non zero vector, the distribution is made to be
uniform from this given viewing direction.
In all cases, NumOfPts is an upper bound the the real number of
distributed points, which will be in the same order.
See also ADAPISO, COVERISO, FFPTDIST.
Example:
Pts1 = CoverPt( solid1, 1000, vector( 0, 0, 0 ) );
Pts2 = CoverPt( solid1, 1000, vector( 0, 0, 1 ) );
Computes two uniform distributions of 1000 points on Solid1. Pts1
is uniform in three space, while Pts2 which is viewed uniform from
the Z direction.
$
CRAISE
CurveType CRAISE( CurveType Curve, NumericType NewOrder )
Raise Curve to the NewOrder Order specified.
Example:
Crv = cbezier( list( ctlpt( E2, -0.7, 0.3 ),
ctlpt( E2, 0.0, 1.0 ),
ctlpt( E2, 0.7, 0.0 ) ) );
Crv2 = CRAISE( Crv, 5 );
raises the 90 degrees corner Bezier curve Crv to be a quartic.
$
CREFINE
CurveType CREFINE( CurveType Curve, NumericType Replace, ListType KnotList )
Provides the ability to Replace a knot vector of Curve, or refine
it. KnotList is a list of knots to refine Curve at. All knots
should be contained in the parametric domain of the Curve. If the knot
vector is replaced, the length of KnotList should be identical to the
length of the original knot vector of the Curve. If Curve is a
Bezier curve, it is automatically promoted to be a Bspline curve.
Example:
Crv2 = CREFINE( Crv, FALSE, list( 0.25, 0.5, 0.75 ) );
refines Crv and adds three new knots at 0.25, 0.5, and 0.75.
$
CREGION
CurveType CREGION( CurveType Curve, NumericType MinParam,
NumericType MaxParam )
Extracts a region from Curve between MinParam and MaxParam.
Both MinParam and MaxParam should be contained in the
parametric domain of the Curve.
Example:
SubCrv = CREGION( Crv, 0.3, 0.6 );
extracts the region from Crv from the parameter value 0.3 to the
parameter value 0.6.
$
CREPARAM
CurveType CREPARAM( CurveType Curve, NumericType MinParam,
NumericType MaxParam )
Reparametrize Curve over a new domain from MinParam to MaxParam.
This operation does not affect the geometry of the curve and only affine
transforms its knot vector. A Bezier curve will automatically be promoted
into a Bspline curve by this function.
Example:
arc1 = arc( vector( 0.0, 0.0, 0.0 ),
vector( 0.5, 2.0, 0.0 ),
vector( 1.0, 0.0, 0.0 ) );
crv1 = arc( vector( 1.0, 0.0, 0.75 ),
vector( 0.75, 0.0, 0.7 ),
vector( 0.5, 0.0, 0.85 ) ) +
arc( vector( 0.5, 0.0, 0.75 ),
vector( 0.75, 0.0, 0.8 ),
vector( 1.0, 0.0, 0.65 ) );
arc1 = CREPARAM( arc1, 0, 10 );
crv1 = CREPARAM( crv1, 0, 10 );
Sets the domain of the given two curves to be from zero to ten. The
Bezier curve arc1 is promoted to a Bspline curve.
$
CROSSEC
PolygonType CROSSEC( PolygonType Object )
This feature is NOT implemented.
$
CRV2TANS
ListType CRV2TANS( CurveType Crv, NumericType FineNess )
Computes all the bi-tangents of Crv, in the XY plane. That
is, all lines that are tangent to Crv at two different locations.
Returned is a list of points with X and Y coefficients representinf the
parametric locations on Crv of the bi-tangent. FineNess
controls the numerical accuracy of the computation. A value of 10 will
provide a good start and the larger this number is, the better the
accuracy will be. See also CRVPTTAN and CNVXHULL.
Example:
Tans = nil();
Crv2Tns = Crv2Tans( Crv, 10 );
for ( i = 1, 1, sizeof( Crv2Tns ),
pt = nth( Crv2Tns, i ):
snoc( ceval( Crv, coord( pt, 0 ) ) +
ceval( Crv, coord( pt, 1 ) ), Tans ) );
finds the bi-tangents of Crv and convert them to a set of line
segments.
$
CRVLNDST
NumericType CRVLNDST( CurveType Crv, PointType PtOnLine, VectorType LnDir,
NumericType IsMinDist, NumericType Epsilon )
or
ListType CRVLNDST( CurveType Crv, PointType PtOnLine, VectorType LnDir,
NumericType IsMinDist, NumericType Epsilon )
Computes the closest (if IsMinDist is TRUE, farthest if FALSE) point
on Curve to the line specified by PtOnLine and LnDir as a
point on the line and a line direction.
Since this operation is partially numeric, Epsilon is used to set
the needed accuracy. It returns the parameter value of the location on
Crv closest to the line.
If, however, Epsilon is negative, -Epsilon is used instead,
and all local extrema in the distance function are returned as a list
(both minima and maxima).
If the line and the curve intersect, the point of intersection is
returned as the minimum.
Example:
Param = CRVLNDST( Crv, linePt, lineVec, TRUE, 0.001 );
finds the closest point on Crv to the line defined by linePt
and lineVec.
$
CRVPTDST
NumericType CRVPTDST( CurveType Crv, PointType Point, NumericType IsMinDist,
NumericType Epsilon )
or
ListType CRVPTDST( CurveType Crv, PointType Point, NumericType IsMinDist,
NumericType Epsilon )
Computes the closest (if IsMinDist is TRUE, farthest if FALSE) point
on Crv to Point.
Since this operation is partially numeric, Epsilon is used to set
the needed accuracy. It returns the parameter value of the location on
Crv closest to Point.
If, however, Epsilon is negative, -Epsilon is used instead,
and all local extrema in the distance function are returned as a list
(both minima and maxima).
Example:
Param = CRVPTDST( Crv, Pt, FALSE, 0.0001 );
finds the farthest point on Crv from point Pt.
$
CRVPTTAN
ListType CRVPTTAN( CurveType Crv, PointType Pt, NumericType FineNess )
Computes all the tangents to Crv that goes through point Pt,
all in the XY plane. Returned is a list of points with X and Y
coefficients representinf the parametric locations on Crv of the
bi-tangent. FineNess controls the numerical accuracy of the
computation. A value of 0.01 will provide a good start and the smaller
this number is, the better the accuracy will be. See also CRV2TANS and
CNVXHULL.
Example:
Tans = nil();
Pt = point( 2, 0, 0 );
CrvPtTns = CrvPtTan( Crv, Pt, 0.01 );
for ( i = 1, 1, sizeof( CrvPtTns ),
snoc( ceval( Crv, nth( CrvPtTns, i ) ) + coerce( Pt, e3 ), Tans ) );
finds the tangents of Crv through Pt and convert them to a
set of line segments.
$
CSURFACE
CurveType CSURFACE( SurfaceType Srf, ConstantType Direction,
NumericType Param )
or
CurveType CSURFACE( TriSrfType Srf, ConstantType Direction,
NumericType Param )
Extract an isoparametric curve out of Srf in the specified
Direction (ROW or COL (or DEPTH for triangular surface) at the
specified parameter value Param.
Param must be contained in the parametric domain of Srf in
Direction direction.
The returned curve is in the surface Srf.
Example:
Crv = CSURFACE( Srf, COL, 0.45 );
extracts an isoparametric curve in the COLumn direction at the parameter
value of 0.15 from surface Srf. See also CMESH, COMPOSE.
$
CTANGENT
VectorType CTANGENT( CurveType Curve, NumericType Param )
Computes the tangent vector to Curve at the parameter value Param.
The returned vector has a unit length.
Example:
Tang = CTANGENT( Crv, 0.5 );
computes the tangent vector to Crv at the parameter value of 0.5.
$
CTLPT
CPt = CTLPT( ConstantType PtType, NumericType Coord1, ... )
Constructs a single control point to be used in the construction of curves
and surfaces. Points can have from one to five dimensions, and may be
either Euclidean or Projective (rational). Points' type is set via the
constants E1 to E5 and P1 to P5. The coordinates of the point are specified
in order, weight is first if rational.
Examples:
CPt1 = CTLPT( E3, 0.0, 0.0, 0.0 );
CPt2 = CTLPT( P2, 0.707, 0.707, 0.707 );
constructs an E3 point at the origin and a P2 rational point with
a weight of 0.707. The Projective Pi points are specified as
CTLPT(Pn, W, W X1, ... , W Xn).
$
CTRIMSRF
ListType CTRIMSRF( TrimSrfType TSrf, NumericType Parametric )
Extract the trimming curves of a trimmed surface TSrf.
If Parametric is not zero, then the trimming curves are extracted
as parametric space curves of TSrf. Otherwise, the trimming curves
are evaluated into Euclidean space as curves on surface TSrf.
Example:
TrimCrvs = CTRIMSRF( TrimSrf, FALSE );
extracts the trimming curves of TrimSrf as Euclidean curves on
TrimSrf.
$
CYLIN
PolylineType CYLIN( VectorType Center, VectorType Direction,
NumericType Radius )
Creates a CYLINder geometric object, defined by Center as center of
the base of the CYLINder, Direction as the CYLINder's axis and height,
and Radius as the radius of the base of the CYLINder.
See RESOLUTION for the accuracy of the CYLINder approximation as a
polygonal model. See IRITSTATE's "PrimRatSrfs" and "PrimRatSrfs" state
variables.
Example:
Cylinder1 = CYLIN( vector( 0, 0, 0 ), vector( 1, 0, 0 ), 10 );
constructs a cylinder along the X axis from the origin to X = 10.
$
CZEROS
ListType CZEROS( CurveType Crv, NumericType Epsilon, NumericType Axis )
Computes the zero set of the given Crv in the given axis (1 for X,
2 for Y, 3 for Z). Since this computation is numeric, an Epsilon is
also required to specify the desired tolerance. It returns a list of
all the parameter values (NumericType) the curve is zero.
Example:
xzeros = CZEROS( cb, 0.001, 1 );
pt_xzeros = nil();
pt = nil();
for ( i = 1, 1, sizeof( xzeros ),
pt = ceval( cb, nth( xzeros, i ) ):
snoc( pt, pt_xzeros )
);
interact( list( axes, cb, pt_xzeros ), 0 );
Computes the X zero set of curve cb with error tolerance
of 0.001. This set is then scanned in a loop and evaluated to
the curve's locations, which are then displayed.
See also CINFLECT.
$
EVOLUTE
CurveType EVOLUTE( CurveType Curve )
or
SurfaceType EVOLUTE( SurfaceType Curve )
Computes the evolute of a curve or a surface.
For curves, the evolute is defined as,
N(t)
E(t) = C(t) + ----
k(t)
where N(t) is the unit normal of C(t) and k(t) is its curvature.
E(t) is computed symbolically as the symbolic sum of C(t) and
N(t) / k(t).
For surfaces, this function computes the mean evulate which is equal to,
n(u, v)
E(u, v) = S(u, v) + ---------
2 H(u, v)
where n(u, v) is the unit normal of S(u, v) and H(u, v) is its mean
curvature. E(u, v) is computed symbolically.
The result of this symbolic computation is exact (upto machine precision)
unlike a similar operations that are only approximated, like the OFFSET or
the AOFFSET.
Example:
crv = cbspline( 3,
list( ctlpt( E3, -1.0, 0.1, 0.2 ),
ctlpt( E3, -0.1, 1.0, 0.1 ),
ctlpt( E3, 0.1, 0.1, 1.0 ),
ctlpt( E3, 1.0, 0.1, 0.1 ),
ctlpt( E3, 0.1, 1.0, 0.2 ) ),
list( KV_OPEN ) );
cev = EVOLUTE( Crv );
$
EXTRUDE
PolygonType EXTRUDE( PolygonType Object, VectorType Dir )
or
SurfaceType EXTRUDE( CurveType Object, VectorType Dir )
Creates an extrusion of the given Object. If Object is a
PolygonObject, its first polygon is used as the base for the extrusion in
Dir direction, and a closed PolygonObject is constructed. If Object
is a CurveType, an extrusion surface is constructed instead, which is not
a closed object (the two bases of the extrusion are excluded, and the curve
may be open by itself).
Direction Dir cannot be coplanar with the polygon plane. The curve
may be nonplanar.
Example:
Cross = cbspline( 3,
list( ctlpt( E2, -0.018, 0.001 ),
ctlpt( E2, 0.018, 0.001 ),
ctlpt( E2, 0.019, 0.002 ),
ctlpt( E2, 0.018, 0.004 ),
ctlpt( E2, -0.018, 0.004 ),
ctlpt( E2, -0.019, 0.001 ) ),
list( KV_OPEN ) );
Cross = Cross + -Cross * scale( vector( 1, -1, 1 ) );
Napkin = EXTRUDE( Cross * scale( vector( 1.6, 1.6, 1.6 ) ),
vector( 0.02, 0.03, 0.2 ) );
constructs a closed cross section Cross by duplicating one half of
it in reverse and merging the two sub-curves. Cross is then used as
the cross-section for the extrusion operation.
$
FFCOMPAT
FFCOMPAT( CurveType Crv1, CurveType Crv2 )
or
FFCOMPAT( SurfaceType Srf1, SurfaceType Srf2 )
Makes the given two curves or surfaces compatible by making them share the
same point type, same curve type, same degree, and the same continuity.
Same point type is gained by promoting a lower dimension into a higher one,
and non-rational to rational points. Bezier curves are promoted to Bspline
curves if necessary, for curve type compatibility. Degree compatibility is
achieved by raising the degree of the lower order curve. Continuity is
achieve by refining both curves to the space with the same (unioned) knot
vector. This function returns nothing and compatibility is made
in place.
Example:
FFCOMPAT( Srf1, Srf2 );
See also CMORPH and SMORPH.
$
FFCTLPTS
ListType FFCTLPTS( CurveType Crv );
or
ListType FFCTLPTS( SurfaceType Srf );
or
ListType FFCTLPTS( TrimSrfType Srf );
or
ListType FFCTLPTS( TrivarType Srf );
Returns all the control points of the given freeform in a single list.
See Also FFKNTVEC, FFMSIZE, FFORDER, FFPTTYPE.
Example:
Ctls = FFCTLPTS( Srf1 );
$
FFEXTREME
CtlPtType FFEXTREME( CurveType Crv, NumericType Minimum )
or
CtlPtType FFEXTREME( SurfaceType Srf, NumericType Minimum )
Computes a bound on the extreme values a curves Crv or surface
Srf can assume. Returned control points provides a bound on the
minimum (maximum) values that can be assumed if Minimum is TRUE
(FALSE).
Example:
Bound = FFEXTREME( Srf, false );
Computes a bound on the maximal values Srf can assume.
$
FFKNTVEC
ListType FFKNTVEC( CurveType Crv )
or
ListType FFKNTVEC( SurfaceType Srf )
or
ListType FFKNTVEC( TrimSrfType TrimSrf )
or
ListType FFKNTVEC( TrivarType Trivar )
or
ListType FFKNTVEC( TriSrfType TriSrf )
Returns all the knot vector(s) of the given freeform in a list of knot
vector(s). See Also FFCTLPTS, FFMSIZE, FFORDER, FFPTTYPE.
Example:
KVs = FFKNTVEC( Srf1 );
$
FFMATCH
FFMATCH( CurveType Crv1, CurveType Crv2, NumericType Reduce,
NumericType Samples, NumericType ReparamOrder,
NumericType Rotate, NumericType NormType )
Computes a reparametrization to Crv2 so it fits Crv1, the best
under some prescribed norm, NormType. Currently the following norms
are valid for NormType
Value Description
1 Suitable for ruled and blended curves, for modeling.
See RULEDSRF.
2 Suitable for metamorphosis of curves. See CMORPH.
3 Distance norm in "walking the dog" notion.
4 Bisector (skeleton) matching norm for two curves.
Whenever negative norms can result (for example, in cases were self
intersection cannot be prevented in ruled surface constructions), one can
allow negativity with no extra penalty by applaying neative NormType.
Use of positive only norms would yield no output at all if no matching
with positive weights can be established whereas allowing negative
norm values would result in the globally optimal result, but with possibly
self intersectiions.
The reparametrization is computed by sampling a fix set of size Samples
off both curves, and fitting a Bspline curve of length Reduce as the
reparametrization curve. Hence, Reduce must be less than or equal to
Samples. The reparametrization curve will have order of
ReparamOrder. If Rotate is TRUE or ON, then attempt is made
to rotate the reparametrization of the curves. Rotation can be used on
closed curves only.
See RULEDSRF and CMORPH for examples.
$
FFMERGE
CurveType FFMERGE( ListType E1Curves, NumericType PointType )@
or
SurfaceType FFMERGE( ListType E1Surfaces, NumericType PointType )
Merges the scalar curves in the list of curves E1Curves or list
of surfaces E1Surfaces to one vector curve/surface of point type
PointType.
Example:
Srf = FFMERGE( list( SrfW, SrfX, SrfY ), P2 );
merges three scalar surfaces into a single surface with point type P2.
See also FFSPLIT, FFPTTYPE.
$
FFMSIZE
ListType FFMSIZE( CurveType Crv )
or
ListType FFMSIZE( SurfaceType Srf )
or
ListType FFMSIZE( TrimSrfType TrimSrf )
or
ListType FFMSIZE( TrivarType Trivar )
or
ListType FFMSIZE( TriSrfType TriSrf )
Returns the size of the control mesh/polygon of the given freeform in a list.
See Also FFCTLPTS, FFKNTVEC, FFORDER, FFPTTYPE.
Example:
MSizes = FFMSIZE( Srf1 );
$
FFORDER
ListType FFORDER( CurveType Crv )
or
ListType FFORDER( SurfaceType Srf )
or
ListType FFORDER( TrimSrfType TrimSrf )
or
ListType FFORDER( TrivarType Trivar )
or
ListType FFORDER( TriSrfType TriSrf )
Return all the orders of the given freeform in a single list.
See Also FFCTLPTS, FFKNTVEC, FFMSIZE, FFPTTYPE.
Example:
Orders = FFORDER( Srf1 );
$
FFPOLES
NumericType FFPOLES( CurveType Crv );
or
NumericType FFPOLES( SurfaceType Srf );
or
NumericType FFPOLES( TrimSrfType TrimSrf );
or
NumericType FFPOLES( TrivarType Trivar );
or
NumericType FFPOLES( TriSrfType TriSrf );
Returns TRUE in the given freeform has poles, FALSE otherwise.
Poles are zeros in the weights of rational functions.
Example:
HasPoles = FFPOLES( Srf1 );
$
FFPTDIST
ListType FFPTDIST( CurveType Crv, NumericType Euclid, NumericType NumOfPts )
or
ListType FFPTDIST( SurfaceType Srf, NumericType Euclid, NumericType NumOfPts )
Computes a uniform point distribution for a Crv or Srf. If
Euclid is FALSE the distribution is selected to be uniform in the
parametric space, otherwise if TRUE, the distribution is made uniform in the
Euclidean space. NumOfPts sets the number of points in the distribution.
The returned list of points prescribes parameters values in the freeforms.
For Crv, the returned list is a list of reals, in the parameter space
of Crv. For Srf, the returned list is a list of points, whose
X and Y coefficients holds the U and V parameters of Srf.
See also COVERPT.
Example:
c1 = cbezier( list( ctlpt( E2, -1.0, 0.0 ),
ctlpt( E2, -1.0, 0.1 ),
ctlpt( E2, -0.9, -0.1 ),
ctlpt( E2, 0.9, 0.0 ) ) );
color( c1, magenta );
pts = FFPTDIST( c1, false, 300 );
e2pts = nil();
for ( i = 1, 10, sizeof( pts ),
pt = ceval( c1, coord( nth( pts, i ), 0 ) ):
snoc( pt, e2pts )
);
interact( list( e2pts, c1 ) );
pts = FFPTDIST( c1, true, 300 );
e2pts = nil();
for ( i = 1, 10, sizeof( pts ),
pt = ceval( c1, coord( nth( pts, i ), 0 ) ):
snoc( pt, e2pts )
);
interact( list( e2pts, c1 ) );
Computes the distribution of 100 points in curve c1 which has
highly nonuniform speed characteristics. Two distributions are computed,
one to be uniform in the parametric space and one uniform in the Euclidean
space.
$
FFPTTYPE
NumericType FFPTTYPE( CurveType Crv )
or
NumericType FFPTTYPE( SurfaceType Srf )
or
NumericType FFPTTYPE( TrimSrfType Srf )
or
NumericType FFPTTYPE( TrivarType )
Returns the point type (E2, P4 etc.) of the given freeform.
See Also FFCTLPTS, FFKNTVEC, FFMSIZE, FFORDER.
$
FFSPLIT
ListType FFSPLIT( CurveType Crv )
or
ListType FFSPLIT( SurfaceType Srf )
Splits the given curve Crv or surface Srf into its scalar
components that are returned as a list of curves/surfaces.
Example:
E1Srfs = FFSPLIT( circle( vector( 0, 0, 0 ), 1 ) );
splits the circle which is a curve in P3 into four scalar curves (W, X, Y, Z)
that are returned in a single list.
See also FFMERGE, FFPTTYPE.
$
GBOX
PolygonType GBOX( VectorType Point,
VectorType Dx, VectorType Dy, VectorType Dz )
Creates a parallelepiped - Generalized BOX polygonal object, defined by
Point as base position, and Dx, Dy, Dz as 3 3D vectors to define
the 6 faces of this generalized BOX. The regular BOX object is a special case
of GBOX where Dx = vector(Dx, 0, 0), Dy = vector(0, Dy, 0), and
Dz = vector(0, 0, Dz).
Dx, Dy, Dz must all be independent in order to create an
object with positive volume.
Example:
GB = GBOX( vector( 0.0, -0.35, 0.63 ), vector( 0.5, 0.0, 0.5 ),
vector( -0.5, 0.0, 0.5 ),
vector( 0.0, 0.7, 0.0 ) );
$
GETATTR
AnyType GETATTR( AnyType Obj, StringType Name )
Provides a mechanism to fetch an attribute named Name from object
Obj.
Example:
attrib( axes, "test", 15 );
a = GETATTR( axes, "test" );
will set the value of a to be 15.
$
GETLINE
AnyType GETLINE( NumericType RequestedType )
Provides a method to get input from keyboard within functions and
or subroutines. RequestedType can be one of NUMERIC_TYPE,
POINT_TYPE, VECTOR_TYPE, or PLANE_TYPE in which the entered line
will be parsed into one, three, or four numeric values (sperated by
either spaces or commas) and the proper object will be created and
returned. In any other case, including failure to parse the numeric
input, a STRING_TYPE object will be constructed from the entered line.
Example:
Pt = GETLINE( point_type );
to read one point (three numeric values) from stdin.
$
GPOLYGON
PolygonType GPOLYGON( GeometryTreeType Object, NumericType Normals )
Approximates all Surface(s)/Trimmed surface(s)/Trivariate(s) in Object
with polygons using the RESOLUTION and FLAT4PLY variables. The larger the
RESOLUTION is, the finer (more polygons) the resulting approximation will be.
FLAT4PLY is a Boolean flag controlling the conversion of an (almost) flat
patch into four (TRUE) or two (FALSE) polygons. Normals are computed to
polygon vertices using surface normals, so Gouraud or Phong shading can be
exploited. It returns a single polygonal object.
If Normals is set, surface normals will be evaluated at the
vertices. Otherwise flat shading and constant normals across polygons are
assumed.
Example:
Polys = GPOLYGON( list( Srf1, Srf2, Srf3 ), off );
Converts to polygons the three surfaces Srf1, Srf2, and Srf3
with no normals.
$
GPOLYLINE
PolylineType GPOLYLINE( GeometryTreeType Object, NumericType Optimal )
Converts all Curves(s), (Trimmed) Surface(s), and Trivariate(s) Object
into polylines using the RESOLUTION variable. The larger the RESOLUTION is,
the finer the resulting approximation will be. It returns a single polyline
object.
If Optimal is false, the points are sampled at equally spaced
interval in the parametric space. If Optimal true, a better, more
expensive computationally, algorithm is used to derive optimal sampling
locations as to minimize the maximal distance between the curve and
piecewise linear approximation (L infinity norm).
Example:
Polys = GPOLYLINE( list( Srf1, Srf2, Srf3, list( Crv1, Crv2, Crv3 ) ),
on );
converts to polylines the three surfaces Srf1, Srf2, and Srf3
and the three curves Crv1, Crv2, and Crv3.
$
HERMITE
SurfaceType HERMITE( CurveType Bndry1, CurveType Bndry2,
CurveType Tan1, CurveType Tan2 )
or
CurveType HERMITE( PointType Bndry1, PointType Bndry2,
VectorType Tan1, VectorType Tan2 )
Constructs a cubic fit between Bndry1 and Bndry2 so that
first derivative continuity constraints, as prescribed by Tan1 at
Bndry1 and Tan2 at Bndry2, are preserved.
Returns either a curve or a surface, according to type of input parameters.
Example:
h00 = HERMITE( point( 0, 0, 0 ),
point( 1, 1, 0 ),
vector( 1, 0, 0 ),
vector( 1, 0, 0 ) );
Constructs a curve in the shape of the first basis function of the cubic
Hermite basis functions.
$
INSTANCE
InstanceType INSTANCE( GeometryType Geom );
Creates an instance of the geometry prescribed by Geom. The instance
is created with a unit transformation matrix that can be modified using
a product with a matrix object.
The use of instances is advantageous where the same geometry is to be
displayed/processed in several different locations in space. A modification
of the original geometry Geom will affect all instances that reference
it. The reference is by the original object's name. The original
object can be a single object or a whole hierarchy of objects.
Example:
Tea1 = INSTANCE( "Teapot" ) * trans( vector( 10, 0, 0 ) );
Tea2 = INSTANCE( "Teapot" ) * trans( vector( 20, 0, 0 ) );
Tea3 = INSTANCE( "Teapot" ) * trans( vector( 30, 0, 0 ) );
viewobj( list( Teapot, Tea1, Tea2, Tea3 ) );
will display four teapots 10 units apart in X.
As an instance is affected by an application of a transformation to it,
if both the original object and the instance are in the same hierarchy,
the transformation of the hierarchy will affect the instance twice:
chair = chair(); # Creates a chair object using chair function
chairInst = INSTANCE( "chair" ) * tx( 1 );
Hier = list( chair, chairInst ) * ty( 1 );
will, in Hier, translate chair to (0, 1, 0) but will translate
chairInst to (1, 2, 0) or translate twice in y (once chair and once
chairInst).
$
LOFFSET
CurveType LOFFSET( CurveType Crv, NumericType OffsetDistance,
NumericType NumOfSamples, NumericType NumOfDOF,
NumericType Order )
Approximate an offset of OffsetDistance by sampling NumOfSamples
samples along the offset curve and least square fitting them using a Bspline
curve of order Order and NumOfDOF control points.
Example:
OffCrv1 = LOFFSET( Crv, -0.4, 100, 10, 4 );
See also OFFSET, AOFFSET, and MOFFSET.
$
MERGEPOLY
PolygonType MERGEPOLY( ListType PolyList )
Merges a set of polygonal objects in PolyList list to a single polygonal
object. All elements in ObjectList must be of PolygonType type. This
function performs the same operation as the overloaded ^ operator
would, but might be more convenient to use under some circumstances.
Example:
Vrtx1 = vector( -3, -2, -1 );
Vrtx2 = vector( 3, -2, -1 );
Vrtx3 = vector( 3, 2, -1 );
Vrtx4 = vector( -3, 2, -1 );
Poly1 = poly( list( Vrtx1, Vrtx2, Vrtx3, Vrtx4 ), false );
Vrtx1 = vector( -3, 2, 1 );
Vrtx2 = vector( 3, 2, 1 );
Vrtx3 = vector( 3, -2, 1 );
Vrtx4 = vector( -3, -2, 1 );
Poly2 = poly( list( Vrtx1, Vrtx2, Vrtx3, Vrtx4 ), false );
Vrtx1 = vector( -3, -2, 1 );
Vrtx2 = vector( 3, -2, 1 );
Vrtx3 = vector( 3, -2, -1 );
Vrtx4 = vector( -3, -2, -1 );
Poly3 = poly( list( Vrtx1, Vrtx2, Vrtx3, Vrtx4 ), false );
PolyObj = MERGEPOLY( list( Poly1, Poly2, Poly3 ) );
$
MOFFSET
CurveType MOFFSET( CurveType Crv, NumericType OffsetDistance,
NumericType AngularError )
Computes an offset of OffsetDistance with globally bounded error
(controlled by AngularError). The smaller AngularError is, the better
the approximation to the offset. The bounded error is achieved by adaptive
refinement of the Crv.
The offset is computed via matching of the tangent fields of the
given curve Crv and an arc spanning the same angular domain.
Further, AngularError measures the angular deviation allowed
between the two tangent fields.
Example:
OffCrv1 = MOFFSET( Crv, -0.4, 10 );
OffCrv2 = MOFFSET( Crv, -0.4, 5 );
computes an offset approximation to Crv with OffsetDistance of
-0.4 and AngularError of 10 and 5 degrees, respectively.
See also OFFSET, AOFFSET, LOFFSET, and FFMATCH.
$
MOMENT
PointType MOMENT( CurveType Crv, 0 );
or
VectorType MOMENT( CurveType Crv, 1 );
Approximates the zero and first moment of curve Crv.
Example:
a = circle( vector( 0, 0, 0 ), 1 );
a = cregion( a, 0, 1 );
p = moment( a, 0 );
v = moment( a, 1 );
view(list(a, p, v), on);
a = cregion( a, 0, 1 ) * rz( 45 );
p = moment( a, 0 );
v = moment( a, 1 );
view(list(a, p, v), on);
computes and displays the zero and first moment of a quarter of a circle
in two orientations.
$
MRCHCUBE
PolygonType MRCHCUBE( ListType VolumeSpec,
PointType CubeDim,
NumericType SkipFactor,
NumericType IsoVal )
Applies (a variation of) the marching cubes algorithm (see
W. E.Lorensen and H. E.Cline. "Marching Cubes: A High Resolution 3D
Surface Construction Algorithm." Computer Graphics (SIGGRAPH '87
Proceedings), Vol. 21, No. 4, pp 163-169, July 1987.) to the given
volumetric data set or trivariate. VolumeSpec can be a list of
three or five objects as follows:
3 a triplet of the form
(TrivarType TV, NumericType Axis, NumericType TVNormal)
5 a 5-tuple of the form
(StringType FileName, NumericType DataType, NumericType Width,
NumericType Height, NumericType Depth )
In the first case, the trivariate TV is iso surface contoured at level
IsoVal along the prescribed Axis (Note a trivariate need not
be a scalar function, while Marching Cubes assumes a scalar function).
If TVNormals is not zero, much more accurate normals are derived using
the trivariate function which is also slower. Otherwise, first order
differencing on the cubes is employed for normal estimation.
In the second case, the volume file prescribed by FileName is
loaded and iso surface contoured. The file is assumed to hold
Width * Height * Depth (Width first, Depth order last) scalar
numeric values of type DataType:
1 Regular ascii (separated by while spaces)
2 Two bytes short integer.
3 Four bytes long integer.
4 One byte (char) integer.
5 Four bytes float.
6 Eight bytes double.
Beware of the little vs big Endian problem! We assume here you read in the
volume in the same machine type this file was writeen with.
CubeDim allows the user to prescribe the real cell size (not
necessarily cubical. SkipFactor allow the skipping ofdata in large
data sets. SkipFactor = 1 skips nothing. SkipFactor = 2, skips
every other scalar value, reducing in half all dimensions, etc.
Last but not least, IsoVal sets the iso surface level.
See also COVERISO, TVLOAD, and TMORPH.
Example:
IsoSrf = MRCHCUBE( list( ThreeCyls, 1, TRUE ), point( 1, 1, 1 ), 1, 0.12 );
Iso surface contour the X axis of trivariate ThreeCyls and use
the trivariate to get better normal's estimations. Cell size is unit
cube like, no skipping of data and iso surface level is 0.12.
$
NIL
ListType NIL()
Creates an empty list so data can be accumulated in it.
See CINFLECT or CZEROS for examples. See also LIST and SNOC.
$
OFFSET
PolygonType OFFSET( PolygonType Poly, NumericType OffsetDistance,
NumericType Tolerance, NumericType BezInterp )
or
CurveType OFFSET( CurveType Crv, NumericType OffsetDistance,
NumericType Tolerance, NumericType BezInterp )
or
SurfaceType OFFSET( SurfaceType Srf, NumericType OffsetDistance,
NumericType Tolerance, NumericType BezInterp )
or
TrimSrfType OFFSET( TrimSrfType TrimSrf, NumericType OffsetDistance,
NumericType Tolerance, NumericType BezInterp )
Offsets Poly, Crv, Srf or a TrimSrf, by translating
all the (control) points in the direction of the normal of the poly/curve
or of the (trimmed) surface by an OffsetDistance amount.
For a Poly object, result is an offset of the original polygon/line
in the XY plane and is exact. Hence, Tolerance and BezInterp
parameters are ignored. Result is a mitter joint based offset.
Otherwise, each control point has a node parameter value associated with
it, which is used to compute the normal. The returned curve or surface only
approximates the real offset. If the resulting approximation does not satisfy
the accuracy required by Tolerance, Crv or Srf or
TrimSrf is subdivided and an offset approximation fit is computed to
the two halfs. For curves, one can request a Bezier interpolation scheme in
the offset approximation by setting BezInterp. The BezInterp is
not supported yet for (trimmed) surfaces.
Negative OffsetDistance denotes offset in the reversed direction of the
normal.
Example:
OffCrv = OFFSET( Crv, -0.4, 0.1, off );
offsets Crv by the amount of -0.4 in the reversed normal direction,
Tolerance of 0.1 and no Bezier interpolation.
See also AOFFSET and LOFFSET.
$
ORTHOTOMC
CurveType ORTHOTOMC( CurveType Crv, PointType Pt, NumericType K )
or,
SurfaceType ORTHOTOMC( SurfaceType Srf, PointType Pt, NumericType K )
Computes the K-orthotomic of freeform curves and surfaces. See
Fundamentals of Computer Aided Geometric Design, by J. Hoschek and and
D. Lasser.
Example:
pt = point( 0, 0.35, 0 );
crv = cbezier( list( ctlpt( E2, -0.8, -0.6 ),
ctlpt( E2, -0.3, -0.2 ),
ctlpt( E2, 0.0, 0.0 ),
ctlpt( E2, 0.8, -0.6 ) ) );
Orth = ORTHOTOMC( crv, pt, 2 );
interact( list( Orth, crv, pt ) * tx( 0.5 ) ) );
computes the orthotomic of a cubic Bezier curve that has an inflection
point. Note inflection points are reduced to cusps in the orthotomic
result.
$
PCIRCLE
CurveType PCIRCLE( VectorType Center, NumericType Radius )
Same as CIRCLE but approximates the circle as a polynomial curve.
See CIRCLE.
$
PDECIMATE
PolygonType PDECIMATE( PolygonType Obj, NumericType Dist,
NumericType NumPasses, NumericType DecimRatio,
NumericType MinAspRatio )
Given a polygonal model, Obj, decimate and merge polygons, effectively
reducing the size of the data subject to a maximal deviation distance
Dist, number of reduction passes NumPasses, maximal number
of DecimRatio of reductions per vertex, and aspect ratio of the
triangulation, MinAspRatio.
Example:
gcross = cbspline( 3,
list( ctlpt( E3, 0.3, 0.0, 0.0 ),
ctlpt( E3, 0.1, 0.0, 0.1 ),
ctlpt( E3, 0.1, 0.0, 0.4 ),
ctlpt( E3, 0.5, 0.0, 0.5 ),
ctlpt( E3, 0.6, 0.0, 0.8 ) ),
list( KV_OPEN ) );
resolution = 30;
glass = surfprev( gcross );
pglass = gpolygon( glass, false );
dglass = PDECIMATE( pglass, 0.005, 3, 1, 0 );
creates a surface of a glass, approximate it with polygons and then
decimates the latter.
$
PDOMAIN
ListType PDOMAIN( CurveType Crv )
or
ListType PDOMAIN( SurfaceType Srf )
or
ListType PDOMAIN( TriSrfType Srf )
or
ListType PDOMAIN( TrimSrfType TrimSrf )
or
ListType PDOMAIN( TrivarType TV )
Returns the parametric domain of the curve (TMin, TMax) or of a (trimmed)
surface (UMin, UMax, VMin, VMax) or of a triangular surface or trivariate
function (UMin, UMax, VMin, VMax, WMin, WMax) as a list object.
Example:
circ_domain = PDOMAIN( circle( vector( 0.0, 0.0, 0.0 ), 1.0 ) );
$
PLN3PTS
PlaneType PLN3PTS( PointType Pt1, PointType Pt2, PointType Pt3 )
Computes a plane out of three points.
Example:
Pl1 = PLN3PTS( point( 0, 0, 0 ), point( 0, 1, 0 ), point( 1, 0, 0 ) );
$
POLY
PolygonType POLY( ListType VrtxList, NumericType IsPolyline )
Creates a single polygon/polyline (and therefore open) object, defined by
the vertices in VrtxList (see LIST). All elements in VrtxList
must be of VectorType type. If IsPolyline, a polyline is created,
otherwise a polygon.
Example:
V1 = vector( 0.0, 0.0, 0.0 );
V2 = vector( 0.3, 0.0, 0.0 );
V3 = vector( 0.3, 0.0, 0.1 );
V4 = vector( 0.2, 0.0, 0.1 );
V5 = vector( 0.2, 0.0, 0.5 );
V6 = vector( 0.3, 0.0, 0.5 );
V7 = vector( 0.3, 0.0, 0.6 );
V8 = vector( 0.0, 0.0, 0.6 );
V9 = vector( 0.0, 0.0, 0.5 );
V10 = vector( 0.1, 0.0, 0.5 );
V11 = vector( 0.1, 0.0, 0.1 );
V12 = vector( 0.0, 0.0, 0.1 );
I = POLY( list( V1, V2, V3, V4, V5, V6, V7, V8, V9, V10, V11, V12 ),
FALSE );
constructs an object with a single polygon in the shape of the letter I.
$
PRISA
ListType PRISA( SurfaceType Srfs, NumericType SamplesPerCurve,
NumericType Epsilon, ConstantType Dir, VectorType Space )
Computes a layout (prisa) of the given surface(s) Srfs, and returns
a list of surface objects representing the layout.
The surface is approximated to within Epsilon in direction Dir
into a set of ruled surfaces and then developable surfaces that are laid out
flat onto the XY plane. If Epsilon is negative, the piecewise ruled
surface approximation in 3-space is returned.
SamplesPerCurve controls the piecewise linear approximation of the
boundary of the ruled/developable surfaces. Space is a vector whose
X component controls the space between the different surfaces' layout, and
whose Y component controls the space between different layout pieces.
Example:
cross = cbspline( 3,
list( ctlpt( E3, 0.7, 0.0, 0. ),
ctlpt( E3, 0.7, 0.0, 0.06 ),
ctlpt( E3, 0.1, 0.0, 0.1 ),
ctlpt( E3, 0.1, 0.0, 0.6 ),
ctlpt( E3, 0.6, 0.0, 0.6 ),
ctlpt( E3, 0.8, 0.0, 0.8 ),
ctlpt( E3, 0.8, 0.0, 1.4 ),
ctlpt( E3, 0.6, 0.0, 1.6 ) ),
list( KV_OPEN ) );
wglass = surfrev( cross );
wgl_ruled = PRISA( wglass, 6, -0.1, COL, vector( 0, 0.25, 0.0 ) );
wgl_prisa = PRISA( wglass, 6, 0.1, COL, vector( 0, 0.25, 0.0 ) );
Computes a layout of a wine glass in wgl_prisa and a three-dimensional
ruled surface approximation of wglass in wgl_ruled.
$
PT3BARY
VectorType PT3BARY( PointType Pt1, PointType Pt2, PointType Pt3,
PointType InteriorPt )
Computes the barycentric coordinates of InterPt with respect
to the triangle defined by Pt1, Pt2, Pt3. Returned
is a vector of three coefficents, which are the weights of the three
points of the triangle. InteriorPt is assumed to be in the
triangle.
Example:
Coeffs = PT3BARY( point( 0, 0, 0 ),
point( 1, 0, 0 ),
point( 0, 1, 0 ),
point( 0.25, 0.25, 0.0 ) );
$
PTHMSPR
ListType PTHMSPR( NumericType Size )
Computes a fairly uniform distribution of points on a hemisphere.
Size hints on the distance between adjacent placed points.
Example:
Pts = PTHMSPR( 0.1 );
$
PTLNPLN
VectorType PTLNPLN( PointType LineOrig, VectorType LineRay, PlaneType Plane )
Computes the point of intersection of given line LineOrig,
LineRay with plane Plane.
Example:
InterPt = PtLnPln( point( 1, 0, 1 ), vector( 1, 1, 1 ), Plane( 0, 0, 1, 0 ) );
$
PTPTLN
VectorType PTPTLN( PointType Point, PointType LineOrig, VectorType LineRay )
Computes the point on line LineOrig, LineRay that is closest
to point Point. See also DSTPTLN
Example:
ClosestPt = PTPTLN( point( 0, 0, 0 ), point( 1, 1, 0 ), vector( 1, 1, 1 ) );
$
PTSLNLN
ListType PTSLNLN( PointType Line1Orig, VectorType Line1Ray,
PointType Line2Orig, VectorType Line2Ray )
Computes the closest two points on the two lines defined by point
LineiOrig and ray LineiRay. See also DSTLNLN. Returned
is a list object with the two points.
Example:
ClosestPts = PtsLnLn( point( 1, 0, 0 ), vector( 0, 1, 0 ),
point( 0, 1, 0 ), vector( 1, 0, 0 ) );
$
RULEDSRF
SurfaceType RULEDSRF( CurveType Crv1, CurveType Crv2 )
or
PolygonType RULEDSRF( PolygonType Poly1, PolygonType Poly22 )
Constructs a ruled surface between the two curves Crv1 and Crv2
or two polylines Poly1 and Poly2.
The curves do not have to have the same order or type, and will be promoted
to their least common denominator. The polys must have the same number of
points and both must be either polygons or polylines.
Example:
c1 = cbspline( 3,
list( ctlpt(E3, 1.7, 0.0 , 0 ),
ctlpt(E3, 0.7, 0.7 , 0 ),
ctlpt(E3, 1.7, 0.3 , 0 ),
ctlpt(E3, 1.5, 0.8 , 0 ),
ctlpt(E3, 1.6, 1.0 , 0 ) ),
list( KV_OPEN ) );
c2 = cbspline( 3,
list( ctlpt(E3, 0.7, 0.0 , 0 ),
ctlpt(E3,-0.7, 0.2 , 0 ),
ctlpt(E3, 0.7, 0.5 , 0 ),
ctlpt(E3,-0.7, 0.7 , 0 ),
ctlpt(E3, 0.7, 1.0 , 0 ) ) ,
list( KV_OPEN ) );
srf1 = RULEDSRF( c1, c2 );
interact( list( c1, c2, srf1 ), on );
c2a = ffmatch( c1, c2, 50, 100, 2, false, 1 );
srf2 = RULEDSRF( c1, c2a );
interact( list( c1, c2, srf2 ), on );
Constructs a planar ruled surface between two curves, c1 and c2.
The naive construction causes self intersection, but by employing FFMATCH
the self intersection can be resloved.
See also FFMATCH.
$
SBEZIER
SurfaceType SBEZIER( ListType CtlMesh )
Creates a Bezier surface using the provided control mesh. CtlMesh is a
list of rows, each of which is a list of control points. All control points
must be of type (E1-E5, P1-P5), or regular PointType defining the surface's
control polygon. Surface's point type will be of a space which is the union
of the spaces of all points.
Example:
Srf = SBEZIER( list ( list( ctlpt( E3, 0.0, 0.0, 1.0 ),
ctlpt( E3, 0.0, 1.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 1.0 ) ),
list( ctlpt( E3, 1.0, 0.0, 0.0 ),
ctlpt( E3, 1.0, 1.0, 2.0 ),
ctlpt( E3, 1.0, 2.0, 0.0 ) ),
list( ctlpt( E3, 2.0, 0.0, 2.0 ),
ctlpt( E3, 2.0, 1.0, 0.0 ),
ctlpt( E3, 2.0, 2.0, 2.0 ) ),
list( ctlpt( E3, 3.0, 0.0, 0.0 ),
ctlpt( E3, 3.0, 1.0, 2.0 ),
ctlpt( E3, 3.0, 2.0, 0.0 ) ),
list( ctlpt( E3, 4.0, 0.0, 1.0 ),
ctlpt( E3, 4.0, 1.0, 0.0 ),
ctlpt( E3, 4.0, 2.0, 1.0 ) ) ) );
$
SBISECTOR
SurfaceType SBISECTOR( SurfaceType Srf, PointType Pt )
Computes the bisector surface of a given surface a a point.
See also CBISECTOR.
Example:
s = ruledSrf( ctlpt( E3, -1.0, -1.0, 0.0 ) + ctlpt( E3, 1.0, -1.0, 0.0 ),
ctlpt( E3, -1.0, 1.0, 0.0 ) + ctlpt( E3, 1.0, 1.0, 0.0 ) );
pt = point( 0.0, 0.0, 1.0 );
bisect = SBISECTOR( s, pt );
interact( list( s, pt, bisect ) );
computes the bisector surface of a plane and a point.
$
SBSPLINE
SurfaceType SBSPLINE( NumericType UOrder, NumericType VOrder,
ListType CtlMesh, ListType KnotVectors )
Creates a Bspline surface from the provided UOrder and VOrder
orders, the control mesh CtlMesh, and the two knot vectors KnotVectors.
CtlMesh is a list of rows, each of which is a list of control points.
All control points must be of point type (E1-E5, P1-P5), or regular
PointType defining the surface's control mesh. Surface's point type will
be of a space which is the union of the spaces of all points.
KnotVectors is a list of two knot vectors. Each knot vector is a
list of NumericType knots of length #CtlPtList plus the Order.
If, however, the length of the knot vector is equal to #CtlPtList +
Order + Order - 1} the curve is assumed periodic.
The knot vector may also be a list of a single constant KV_OPEN or
KV_FLOAT or KV_PERIODIC, in which a uniform knot vector with the
appropriate length and with open, floating or periodic end condition
will be constructed automatically.
Example:
Mesh = list ( list( ctlpt( E3, 0.0, 0.0, 1.0 ),
ctlpt( E3, 0.0, 1.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 1.0 ) ),
list( ctlpt( E3, 1.0, 0.0, 0.0 ),
ctlpt( E3, 1.0, 1.0, 2.0 ),
ctlpt( E3, 1.0, 2.0, 0.0 ) ),
list( ctlpt( E3, 2.0, 0.0, 2.0 ),
ctlpt( E3, 2.0, 1.0, 0.0 ),
ctlpt( E3, 2.0, 2.0, 2.0 ) ),
list( ctlpt( E3, 3.0, 0.0, 0.0 ),
ctlpt( E3, 3.0, 1.0, 2.0 ),
ctlpt( E3, 3.0, 2.0, 0.0 ) ),
list( ctlpt( E3, 4.0, 0.0, 1.0 ),
ctlpt( E3, 4.0, 1.0, 0.0 ),
ctlpt( E3, 4.0, 2.0, 1.0 ) ) );
Srf = SBSPLINE( 3, 3, Mesh, list( list( KV_OPEN ),
list( 3, 3, 3, 4, 5, 6, 6, 6 ) ) );
constructs a bi-quadratic Bspline surface with its first knot vector
having uniform knot spacing with open end conditions.
$
SCRVTR
SurfaceType SCRVTR( SurfaceType Srf, ConstType PtType, ConstType Dir )
Symbolically computes the extreme curvature bound on Srf.
If Dir is either ROW or COL, then the normal curvature square
of Srf in Dir is computed symbolically and returned.
Otherwise, a upper bound on the sum of the squares of the two
principal curvatures is symbolically computed and returned.
Returned value is a surface that can be evaluated to the curvature
bound, given a UV location. The returned surface value is a scalar field
of point type P1 (scalar rational).
However, if PtType is one of E1, P1, E3, P3 the returned surface
is coerced to this given type. If the types are one of E3, P3, then the
Y and Z axes are set to be equivalent to the U and V parametric domains.
This function computes the square of the normal curvature scalar
field for surfaces as (in the U parametric direction, same for V),
2
d S
< n , --- >
2
u du
k (u, v) = ------------
n
dS ds
< -- , -- >
du du
and computes the sum of the squares of the principal curvatures as,
2 2 ( g11 l22 + g22 l11 - 2 g12 l12 )^2 - 2 |G| |L|
k + k = -----------------------------------------------
1 2 |G|^2 ||n||^2
See also CCRVTR.
Example:
cross = cbspline( 3,
list( ctlpt( E2, 0.0, 0.0 ),
ctlpt( E2, 0.8, 0.0 ),
ctlpt( E2, 0.8, 0.2 ),
ctlpt( E2, 0.07, 1.4 ),
ctlpt( E2, -0.07, 1.4 ),
ctlpt( E2, -0.8, 0.2 ),
ctlpt( E2, -0.8, 0.0 ),
ctlpt( E2, 0.0, 0.0 ) ),
list( KV_OPEN ) );
cross = coerce( cross, e3 );
s = sFromCrvs( list( cross,
cross * trans( vector( 0.5, 0, 1 ) ),
cross * trans( vector( 0, 0, 2 ) ) ), 3 );
view( list( s, axes ), on );
UCrvtrZXY = scrvtr( s, E3, row );
VCrvtrZXY = scrvtr( s, E3, col );
UCrvtrXYZ = UCrvtrZXY * rotx( -90 ) * roty( -90 ) * scale( vector( 1, 1, 0.001 ) );
VCrvtrXYZ = VCrvtrZXY * rotx( -90 ) * roty( -90 ) * scale( vector( 1, 1, 10 ) );
color( UCrvtrXYZ, red );
color( VCrvtrXYZ, magenta );
view( list( UCrvtrXYZ, VCrvtrXYZ ), off );
CrvtrZXY = scrvtr( s, E3, off );
CrvtrXYZ = CrvtrZXY * rotx( -90 ) * roty( -90 ) * scale( vector( 1, 1, 0.001 ) );
color( CrvtrXYZ, green );
view( CrvtrXYZ, off );
Computes the sqaure of the normal curvature in the U and V direction, flips
its scalar value from X to Z using rotations and scale the fields to
reasonable values and display them.
Then, display a total bound on the normal curvature as well.
Due to the large degree of the resulting fields be warned that rational
surfaces will compute into large degree curvature bound fields. See also
IRITSTATE("InterpProd", FALSE); for faster symbolic computation.
$
SDERIVE
SurfaceType SDERIVE( SurfaceType Srf, NumericType Dir )
Returns a vector field surface representing the differentiated surface
in the given direction (ROW or COL). Evaluation of the returned surface at
a given parameter value will return a vector tangent to Srf in
Dir at that parameter value.
DuSrf = SDERIVE( Srf, ROW );
DvSrf = SDERIVE( Srf, COL );
Normal = coerce( seval( DuSrf, 0.5, 0.5 ), VECTOR_TYPE ) ^
coerce( seval( DvSrf, 0.5, 0.5 ), VECTOR_TYPE );
computes the two partial derivatives of the surface Srf and computes
its normal as their cross product, at the parametric location (0.5, 0.5).
$
SDIVIDE
SurfaceType SDIVIDE( SurfaceType Srf, ConstantType Direction,
NumericType Param )
or
TrimSrfType SDIVIDE( TrimSrfType Srf, ConstantType Direction,
NumericType Param )
Subdivides a (possibly trimmed) surface into two at the specified parameter
value Param in the specified Direction (ROW or COL). Srf
can be either a Bspline surface in which Param must be conatined in
the parametric domain of the surface, or a Bezier surface in which Param} must be in the range of zero to one.
It returns a list of upto two sub-surfaces. The individual surfaces may be
extracted from the list using the NTH command. If Srf is a trimmed
surface, it can be the case that one of the two subdivided surfaces is
completely trimmed out, and hence only one surface will be returned.
Example:
SrfLst = SDIVIDE( Srf, ROW, 0.5 );
Srf1 = nth( SrfLst, 1 );
Srf2 = nth( SrfLst, 2 );
Subdivides Srf at the parameter value of 0.5 in the ROW direction.
$
SEDITPT
SurfaceType SEDITPT( SurfaceType Srf, CtlPtType CPt, NumericType UIndex,
NumericType VIndex )
Provides a simple mechanism to manually modify a single control point number
UIndex and VIndex (base count is 0) in the control mesh of Srf
by substituting CtlPt instead. CtlPt must have the same point type as
the control points of Srf. Original surface Srf is not modified.
Example:
CPt = ctlpt( E3, 1, 2, 3 );
NewSrf = SEDITPT( Srf, CPt, 0, 0 );
Constructs a NewSrf with the first control point of Srf being
CPt.
$
SEVAL
CtlPtType SEVAL( SurfaceType Srf, NumericType UParam, NumericType VParam )
or
CtlPtType SEVAL( TrimSrfType Srf, NumericType UParam, NumericType VParam )
Evaluates the provided (possibly trimmed) surface Srf at the given
UParam and VParam parameters. Both UParam and VParam
should be contained in the surface parametric domain if Srf is a
Bspline surface, or between zero and one if Srf is a Bezier surface.
The returned control point has the same type as the control points of
Srf.
Example:
CPt = SEVAL( Srf, 0.25, 0.22 );
Evaluates Srf at the parameter values of (0.25, 0.22).
$
SFOCAL
SurfaceType SFOCAL( SurfaceType Srf, NumericType Dir )
Evaluates the focal surface field of surface Srf using the normal
curvature in the isoparametric direction as given by Dir (either
ROW or COL). Note this function is not using the principal curvatures as
is generaly the case for focal surfaces.
Example:
gcross = cbspline( 3,
list( ctlpt( E3, 0.3, 0.0, 0.0 ),
ctlpt( E3, 0.1, 0.0, 0.1 ),
ctlpt( E3, 0.1, 0.0, 0.4 ),
ctlpt( E3, 0.5, 0.0, 0.5 ),
ctlpt( E3, 0.6, 0.0, 0.8 ) ),
list( KV_OPEN ) );
glass = surfprev( gcross );
color( glass, red );
gfocal = SFOCAL(glass, col);
Evaluates the focal surface using the COL isoparametric direction's
normal curvature of the glass surface.
$
SFROMCRVS
SurfaceType SFROMCRVS( ListType CrvList, NumericType OtherOrder )
Constructs a surface by substituting the curves in CrvList as rows
in a control mesh of a surface. Curves in CrvList are made compatible
by promoting Bezier curves to Bsplines if necessary, and raising degree
and refining as required before substituting the control polygons of the
curves as rows in the mesh. The other direction order is set by
OtherOrder, which cannot be larger than the number of curves.
The surface interpolates the first and last curves only.
Example:
Crv1 = cbspline( 3,
list( ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 1.0, 0.0, 0.0 ),
ctlpt( E3, 1.0, 1.0, 0.0 ) ),
list( KV_OPEN ) );
Crv2 = Crv1 * trans( vector( 0.0, 0.0, 1.0 ) );
Crv3 = Crv2 * trans( vector( 0.0, 1.0, 0.0 ) );
Srf = SFROMCRVS( list( Crv1, Crv2, Crv3 ), 3 );
$
SGAUSS
SurfaceType SGAUSS( SurfaceType Srf )
Evaluates the Gaussian curvature field of surface Srf.
Example:
Srf1 = hermite( cbezier( list( ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 0.5, 0.2, 0.0 ),
ctlpt( E3, 1.0, 0.0, 0.0 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 1.0, 0.0 ),
ctlpt( E3, 0.5, 0.8, 0.0 ),
ctlpt( E3, 1.0, 1.0, 0.5 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ) ) ) );
SGauss = SGAUSS( Srf1 );
Evaluates the Gaussian curvaure of Srf1.
$
SINTERP
SurfaceType SINTERP( ListType PtList, NumericType UOrder, NumericType VOrder,
NumericType USize, NumericType VSize,
ConstantType Param)
Computes a Bspline polynomial surface that interpolates or approximates
the rectangular grid or scattered set of points in PtList. The
Bspline surface will have orders UOrder and VOrder and mesh
of size USize by VSize control points. If the data is on a grid,
the knots will be spaced according to Param which can be one of
PARAM_UNIFORM, PARAM_CHORD or PARAM_CENTRIP. The former prescribed a
uniform knot sequence and the latters specified a knot spacing according
to the chord length and a square root of the chord length. Currently only
PARAM_UNIFORM is supported. For scattered point set, the Param
parameter is ignored.
PtList is a list of list of points for girdded data where all
lists should carry the same amount of points in them, defining a
rectangular grid. For scattered data, PtList is a linear list
of points. All points in PtList must be of type (E1-E5, P1-P5)
control point, or regular PointType. If USize and VSize
are equal to the number of points in the grid data set of PtList,
the resulting surface will interpolate the data set. Otherwise,
if USize or VSize is less than the number of points in
the grid of PtList, the point data set will be least square
approximated. In no time can USize or VSize be larger that the
number of points in PtList or lower than UOrder and VOrder,
respectively. If USize or VSize are zero, the grid size
is used, forcing an interpolation of the data set.
If PtList contains a linear list of points, these points are
treated as scattered. Each scattered point is assumed to be holding the
parameteric location to interpolate at as its first two coefficients.
The other coefficients are the interpolation values. In other words,
to interpolated scattered data of type E3, E5 control points in a linear
list must be provided in (u, v, x, y, z) format. Scattered data is
interpolated over a unit square (0 to 1) parameteric domain in both u
and v.
All interior knots will be distinct preserving maximal continuity.
The resulting Bspline surface will have open end conditions.
Example:
pl = nil();
pll = nil();
for ( x = -5, 1, 5,
pl = nil():
for ( y = -5, 1, 5,
snoc( point( x, y, sin( x * Pi / 2 ) * cos( y * Pi / 2 ) ),
pl )
):
snoc( pl, pll ) );
s1 = sinterp( pll, 3, 3, 8, 8, PARAM_UNIFORM );
s2 = sinterp( pll, 3, 3, 11, 11, PARAM_UNIFORM );
Samples an explicit surface sin(x) * cos(y) at a grid of 11 by 11 points,
least square fit with a grid of size of 8 by 8 surface s1, and
interpolate surface s2 using this data set.
$
SMEANSQR
SurfaceType SMEANSQR( SurfaceType Srf )
Evaluates the square of the mean curvature field of surface Srf.
Example:
Srf1 = hermite( cbezier( list( ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 0.5, 0.2, 0.0 ),
ctlpt( E3, 1.0, 0.0, 0.0 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 1.0, 0.0 ),
ctlpt( E3, 0.5, 0.8, 0.0 ),
ctlpt( E3, 1.0, 1.0, 0.5 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ) ) ),
cbezier( list( ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 0.0 ) ) ) );
SMean = SMEANSQR( Srf1 );
Evaluates the square of the mean curvaure of Srf1.
$
SMERGE
SurfaceType SMERGE( SurfaceType Srf1, SurfaceType Srf2,
NumericType Dir, NumericType SameEdge )
Merges two surfaces along the requested direction (ROW or COL). If
SameEdge is non-zero (ON or TRUE), then the common edge is assumed to be
identical and copied only once. Otherwise (OFF or FALSE), a ruled surface
is constructed between the two surfaces along the (not) common edge.
Example:
MergedSrf = SMERGE( Srf1, Srf2, ROW, TRUE );
$
SMORPH
SurfaceType SMORPH( SurfaceType Srf1, SurfaceType Srf2, NumericType Blend )
Creates a new surface which is a convex blend of the two given surfaces.
The two given surfaces must be compatible (see FFCOMPAT) before this blend
is invoked. Very useful if a sequence that "morphs" one surface to another
is to be created.
Example:
for ( i = 0.0, 1.0, 11.0,
Msrf = SMORPH( Srf1, Srf2, i / 11.0 ):
color( Msrf, white ):
attrib( Msrf, "rgb", "255,255,255" ):
attrib( Msrf, "reflect", "0.7" ):
save( "morp1-" + i, Msrf )
);
creates a sequence of 12 surfaces, morphed from Srf1 to Srf2
and saves them in the files "morph-0.dat" to "morph-11.dat".
See also CMORPH and TMORPH.
$
SNORMAL
VectorType SNORMAL( SurfaceType Srf, NumericType UParam, NumericType VParam )
or
VectorType SNORMAL( TrimSrfType Srf, NumericType UParam, NumericType VParam )
Computes the normal vector to (possibly trimmed) surface Srf at the
parameter values UParam and VParam. The returned vector has a
unit length.
Example:
Normal = SNORMAL( Srf, 0.5, 0.5 );
computes the normal to Srf at the parameter values (0.5, 0.5).
See also SNRMLSRF.
$
SNRMLSRF
SurfaceType SNRMLSRF( SurfaceType Srf )
Symbolically computes a vector field surface representing the non-normalized
normals of the given surface. That is the normal surface, evaluated at
(u, v), provides a vector in the direction of the normal of the original
surface at (u, v). The normal surface is computed as the symbolic cross
product of the two surfaces representing the partial derivatives of the
original surface.
Example:
NrmlSrf = SNRMLSRF( Srf );
$
SPHERE
PolygonType SPHERE( VectorType Center, NumericType Radius )
Creates a SPHERE geometric object, defined by Center as the center of
the SPHERE, and with Radius as the radius of the SPHERE.
See RESOLUTION for accuracy of SPHERE approximation as a polygonal model.
See IRITSTATE's "PrimRatSrfs" and "PrimRatSrfs" state variables.
$
SRAISE
SurfaceType SRAISE( SurfaceType Srf, ConstantType Direction,
NumericType NewOrder )
Raises Srf to the specified NewOrder in the specified
Direction.
Example:
Srf = ruledSrf( cbezier( list( ctlpt( E3, -0.5, -0.5, 0.0 ),
ctlpt( E3, 0.5, -0.5, 0.0 ) ) ),
cbezier( list( ctlpt( E3, -0.5, 0.5, 0.0 ),
ctlpt( E3, 0.5, 0.5, 0.0 ) ) ) );
Srf = SRAISE( SRAISE( Srf, ROW, 3 ), COL, 3 );
constructs a bilinear flat ruled surface and raises both its directions to be
a bi-quadratic surface.
$
SREFINE
SurfaceType SREFINE( SurfaceType Srf, ConstantType Direction,
NumericType Replace, ListType KnotList )
Provides the ability to Replace a knot vector of Srf or refine
it in the specified direction Direction (ROW or COL).
KnotList is a list of knots to refine Srf at. All knots should be
contained in the parametric domain of Srf in Direction. If the knot
vector is replaced, the length of KnotList should be identical to the
length of the original knot vector of Srf in Direction. If Srf
is a Bezier surface, it is automatically promoted to be a Bspline surface.
Example:
Srf = SREFINE( SREFINE( Srf,
ROW, FALSE, list( 0.333, 0.667 ) ),
COL, FALSE, list( 0.333, 0.667 ) );
refines Srf in both directions by adding two more knots at 0.333 and
0.667.
$
SREGION
SurfaceType SREGION( SurfaceType Srf, ConstantType Direction,
NumericType MinParam, NumericType MaxParam )
Extracts a region of Srf between MinParam and MaxParam
in the specified Direction. Both MinParam and MaxParam
should be contained in the parametric domain of Srf in Direction.
Example:
Srf = ruledSrf( cbezier( list( ctlpt( E3, -0.5, -0.5, 0.5 ),
ctlpt( E3, 0.0, 0.5, 0.0 ),
ctlpt( E3, 0.5, -0.5, 0.0 ) ) ),
cbezier( list( ctlpt( E3, -0.5, 0.5, 0.0 ),
ctlpt( E3, 0.0, 0.0, 0.0 ),
ctlpt( E3, 0.5, 0.5, 0.5 ) ) ) );
SubSrf = SREGION( Srf, ROW, 0.3, 0.6 );
extracts the region of Srf from the parameter value 0.3 to the
parameter value 0.6 along the ROW direction. the COLumn direction is
extracted as a whole.
$
SREPARAM
SurfaceType SREPARAM( SurfaceType Srf, ConstantType Direction,
NumericType MinParam, NumericType MaxParam )
Reparametrize Srf over a new domain from MinParam to MaxParam,
in the prescribed Direction. This operation does not affect the geometry
of the surface and only affine transforms its knot vectors. A Bezier surface
will automatically be promoted into a Bspline surface by this function.
Example:
srf = sbspline( 2, 4,
list( list( ctlpt( E3, 0.0, 0.0, 1.0 ),
ctlpt( E2, 0.0, 1.0 ),
ctlpt( E3, 0.0, 2.0, 1.0 ) ),
list( ctlpt( E2, 1.0, 0.0 ),
ctlpt( E3, 1.0, 1.0, 2.0 ),
ctlpt( E2, 1.0, 2.0 ) ),
list( ctlpt( E3, 2.0, 0.0, 2.0 ),
ctlpt( E2, 2.0, 1.0 ),
ctlpt( E3, 2.0, 2.0, 2.0 ) ),
list( ctlpt( E2, 3.0, 0.0 ),
ctlpt( E3, 3.0, 1.0, 2.0 ),
ctlpt( E2, 3.0, 2.0 ) ),
list( ctlpt( E3, 4.0, 0.0, 1.0 ),
ctlpt( E2, 4.0, 1.0 ),
ctlpt( E3, 4.0, 2.0, 1.0 ) ) ),
list( list( KV_OPEN ),
list( KV_OPEN ) ) );
srf = sreparam( sreparam( srf, ROW, 0, 1 ), COL, 0, 1 );
Ensures that the Bspline surface is defined over the unit size parametric
domain.
$
SRINTER
PointType SRINTER( SurfaceType Srf, PointType RayOrigin,
VectorType RayDirection )
Computes the first intersection, if any, of the prescribed ray originating
from RayOrigin in direction RayDirection with surface Srf.
Returns the intersection point in the parametric space of Srf with
the U and V coordinates as the X and Y coefficients of the returned value.
The intersection is computed between the ray and a polygonal approximation
of the surface Srf as set via the RESOLUTION variable.
Example:
RayOrigin = point( 2, 0.1, 0.3 );
RayDir = vector( -4, 0, 0 );
RayLine = coerce( RayOrigin, E3 ) + coerce( RayOrigin + RayDir, E3 );
color( RayLine, magenta );
attrib( RayLine, "dwidth", 2 );
resolution = 5;
InterPt = SRINTER( glass, RayOrigin, RayDir );
InterPtE3 = seval( glass, coord( InterPt, 0 ), coord( InterPt, 1 ) );
color( InterPtE3, cyan );
attrib( InterPtE3, "dwidth", 3 );
view( list( InterPtE3, RayLine, glass, axes ), 1 );
resolution = 80;
InterPt = SRINTER( glass, RayOrigin, RayDir );
InterPtE3 = seval( glass, coord( InterPt, 0 ), coord( InterPt, 1 ) );
color( InterPtE3, cyan );
attrib( InterPtE3, "dwidth", 3 );
view( list( InterPtE3, RayLine, glass, axes ), 1 );
A complete example of constructing a ray and intersecting it against a
surface of a glass at two different resolution, resulting in two different
accuracies. See also RESOLUTION.
$
STANGENT
VectorType STANGENT( SurfaceType Srf, ConstantType Direction,
NumericType UParam, NumericType VParam )
or
VectorType STANGENT( TrimSrfType Srf, ConstantType Direction,
NumericType UParam, NumericType VParam )
Computes the tangent vector to (possibly trimmed) surface Srf at the
parameter values UParam and VParam in Direction. The
returned vector has a unit length.
Example:
Tang = STANGENT( Srf, ROW, 0.5, 0.6 );
computes the tangent to Srf in the ROW direction at the parameter
values (0.5, 0.6).
$
STRIMSRF
SurfaceType STRIMSRF( TrimSrfType TSrf )
Extracts the surface of a trimmed surface TSrf.
Example:
Srf = STRIMSRF( TrimSrf );
extracts the surface of TrimSrf.
$
STRIVAR
SurfaceType STRIVAR( TrivarType TV )
Extracts an iso surface from a trivariate function TV.
Example:
TV1 = tbezier( list( list( list( ctlpt( E3, 0.1, 0.0, 0.8 ),
ctlpt( E3, 0.2, 0.1, 2.4 ) ),
list( ctlpt( E3, 0.3, 2.2, 0.2 ),
ctlpt( E3, 0.4, 2.3, 2.0 ) ) ),
list( list( ctlpt( E3, 2.4, 0.8, 0.1 ),
ctlpt( E3, 2.2, 0.7, 2.3 ) ),
list( ctlpt( E3, 2.3, 2.6, 0.5 ),
ctlpt( E3, 2.1, 2.5, 2.7) ) ) ) );
Srf = STRIVAR( TV1, col, 0.4 );
extracts an iso surface of TV1, in the col direction at
parameter value 0.4.
$
SURFPREV
SurfaceType SURFPREV( CurveType Object )
Same as SURFREV but approximates the surface of revolution as a
polynomial surface. Object must be a polynomial curve.
See SURFREV.
$
SURFREV
PolygonType SURFREV( PolygonType Object )
or
SurfaceType SURFREV( CurveType Object )
Creates a surface of revolution by rotating the first polygon/curve of the
given Object, around the Z axis. Use the linear transformation function
to position a surface of revolution in a different orientation.
Example:
VTailAntn = SURFREV( ctlpt( E3, 0.001, 0.0, 1.0 ) +
ctlpt( E3, 0.01, 0.0, 1.0 ) +
ctlpt( E3, 0.01, 0.0, 0.8 ) +
ctlpt( E3, 0.03, 0.0, 0.7 ) +
ctlpt( E3, 0.03, 0.0, 0.3 ) +
ctlpt( E3, 0.001, 0.0, 0.0 ) );
constructs a piecewise linear Bspline curve in the XZ plane and uses it to
construct a surface of revolution by rotating it around the Z axis.
See also SURFPREV.
$
SVISIBLE
ListType SVISIBLE( SurfaceType Srf,
NumericType Resolution,
NumericType ConeSize )
Computes a decomposition of a freeform surface Srf into regions each
visible with a cone visibility of ConeSize degrees from one direction.
In other words all point in one region have angular deviation of their surface
normal of less than ConeSize degrees from the set viewing direction.
Resolution controls the accuracy of the computation; the higher this
value is, more exact the result. 20 is a good starting value. Each returned
region is a trimmed surface that has a "ViewDir" attribute that contains
the viewing direction of this region.
Example:
c1 = cbezier( list( ctlpt( E3, 1.0, 0.0, 0.5 ),
ctlpt( E3, 1.1, 0.0, 0.0 ),
ctlpt( E3, 1.0, 0.0, -0.5 ) ) );
Simp = sregion( surfPRev( c1 ), col, 0.0, 1.0 ) * rz( 45 ) * rx( 90 );
Decomp = SVISIBLE( Simp, 20, 30 * pi / 180 );
SimDecomp = nil();
Mod = 5;
for ( i = 1, 1, sizeof( Decomp ),
o = nth( Decomp, i ):
v = getattr( o, "ViewDir" ):
l = ( ctlpt( E3, 0, 0, 0 ) + coerce( v, e3 ) ) * sc( 1.5 ):
j = floor( ( i - 1 ) / Mod ):
snoc( list( o, Simp, l, axes )
* view_mat * tx( ( i - 1 - j * Mod ) * 2 - 4 )
* ty( -j * 2 ),
SimDecomp ) );
view( SimDecomp, on );
decomposes a given surface Simp into regions of 30 degrees at most,
goes over the decomposed regions and order them five in a row.
$
SWEEPSRF
SurfaceType SWEEPSRF( CurveType CrossSection, CurveType Axis,
CurveType FrameCrv | VectorType FrameVec | ConstType OFF )
Constructs a generalized cylinder surface. This function sweeps a specified
cross-section CrossSection along the provided Axis.
By default, when frame specification is OFF, the orientation
of the cross section is computed using the Axis curve tangent and
normal. However, unlike the Frenet frame, attempt is made to minimize
the normal change, as can happen along inflection points in Axis.
If a VectorType FrameVec is provided as a frame orientation setting,
it is used to fix the binormal direction to this value. In other words, the
orientation frame has a fixed binormal. If a CurveType FrameCrv is
specified as a frame orientation setting, this vector field curve is
evaluated at each placement of the cross-section to yield the needed binormal.
The resulting sweep is only an approximation of the real sweep. The
resulting sweep surface will not be exact, in general.
Refinement of the axis curve at the proper location, where accuracy is
important, should improve the accuracy of the output. The parametric domains
of FrameCrv do not have to match the parametric domain of Axis,
and its parametric domain is automatically made compatible by this function.
Example:
Cross = arc( vector( 0.2, 0.0, 0.0 ),
vector( 0.2, 0.2, 0.0 ),
vector( 0.0, 0.2, 0.0 ) ) +
arc( vector( 0.0, 0.4, 0.0 ),
vector( 0.1, 0.4, 0.0 ),
vector( 0.1, 0.5, 0.0 ) ) +
arc( vector( 0.8, 0.5, 0.0 ),
vector( 0.8, 0.3, 0.0 ),
vector( 1.0, 0.3, 0.0 ) ) +
arc( vector( 1.0, 0.1, 0.0 ),
vector( 0.9, 0.1, 0.0 ),
vector( 0.9, 0.0, 0.0 ) ) +
ctlpt( E2, 0.2, 0.0 );
Axis = arc( vector( -1.0, 0.0, 0.0 ),
vector( 0.0, 0.0, 0.1 ),
vector( 1.0, 0.0, 0.0 ) );
Axis = crefine( Axis, FALSE, list( 0.25, 0.5, 0.75 ) );
Srf1 = SWEEPSRF( Cross, Axis, OFF );
Srf2 = SWEEPSRF( Cross, Axis, vector( 0.0, 1.0, 1.0 ) );
Srf3 = SWEEPSRF( Cross, Axis,
cbezier( list( ctlpt( E3, 1.0, 0.0, 0.0 ),
ctlpt( E3, 0.0, 1.0, 0.0 ),
ctlpt( E3, -1.0, 0.0, 0.0 ) ) ) );
constructs a rounded rectangle cross-section and sweeps it along an arc,
while orienting it several ways. The axis curve Axis is manually
refined to better approximate the requested shape.
See also SWPSCLSRF for sweep with scale.
$
SWPSCLSRF
SurfaceType SWPSCLSRF( CurveType CrossSection, CurveType Axis,
NumericType Scale | CurveType ScaleCrv,
CurveType FrameCrv | VectorType FrameVec
| ConstType OFF,
NumericType ScaleRefine )
Constructs a generalized cylinder surface. This function sweeps a specified
cross-section CrossSection along the provided Axis.
The cross-section may be scaled by a constant value Scale, or scaled
along the Axis parametric direction via a scaling curve ScaleCrv.
By default, when frame specification is OFF, the orientation
of the cross section is computed using the Axis curve tangent and
normal. However, unlike the Frenet frame, attempt is made to minimize
the normal change, as can happen along inflection points in Axis.
If a VectorType FrameVec is provided as a frame orientation setting,
it is used to fix the binormal direction to this value. In other words, the
orientation frame has a fixed binormal. If a CurveType FrameCrv is
specified as a frame orientation setting, this vector field curve is
evaluated at each placement of the cross-section to yield the needed binormal.
ScaleRefine is an integer value to define possible refinement of the
Axis to reflect the information in ScalingCrv. Value of zero will
force no refinement while value of n > 0 will insert n times the
number of control points in ScaleCrv into Axis, better emulating
the scaling requested.
The resulting sweep is only an approximation of the real sweep. The
scaling and axis placement will not be exact, in general.
Manual refinement (in addition to ScaleRefine) of the axis curve at
the proper location, where accuracy is important, should improve the accuracy
of the output. The parametric domains of ScaleCrv and FrameCrv do
not have to match the parametric domain of Axis, and their domains are
made compatible by this function.
Example:
Cross = arc( vector( -0.11, -0.1, 0.0 ),
vector( -0.1, -0.1, 0.0 ),
vector( -0.1, -0.11, 0.0 ) ) +
arc( vector( 0.1, -0.11, 0.0 ),
vector( 0.1, -0.1, 0.0 ),
vector( 0.11, -0.1, 0.0 ) ) +
arc( vector( 0.11, 0.1, 0.0 ),
vector( 0.1, 0.1, 0.0 ),
vector( 0.1, 0.11, 0.0 ) ) +
arc( vector( -0.1, 0.11, 0.0 ),
vector( -0.1, 0.1, 0.0 ),
vector( -0.11, 0.1, 0.0 ) ) +
ctlpt( E2, -0.11, -0.1 );
scaleCrv = cbspline( 3,
list( ctlpt( E2, 0.05, 1.0 ),
ctlpt( E2, 0.1, 0.0 ),
ctlpt( E2, 0.2, 2.0 ),
ctlpt( E2, 0.3, 0.0 ),
ctlpt( E2, 0.4, 2.0 ),
ctlpt( E2, 0.5, 0.0 ),
ctlpt( E2, 0.6, 2.0 ),
ctlpt( E2, 0.7, 0.0 ),
ctlpt( E2, 0.8, 2.0 ),
ctlpt( E2, 0.85, 1.0 ) ),
list( KV_OPEN ) );
Axis = circle( vector( 0, 0, 0 ), 1 );
Frame = circle( vector( 0, 0, 0 ), 1 )
* rotx( 90 ) * trans( vector( 1.5, 0.0, 0.0 ) );
Srf1 = SWPSCLSRF( Cross, Axis, scaleCrv, off, 0 );
Srf2 = SWPSCLSRF( Cross, Axis, scaleCrv, off, 2 );
Srf3 = SWPSCLSRF( Cross, Axis, 1.0, Frame, 0 );
constructs a rounded rectangle cross-section and sweeps it along a circle,
while scaling and orienting in several ways. The axis curve Axis is
automatically refined in Srf2 to better approximate the requested
scaling.
See also SWEEPSRF for sweep with no scale.
$
SYMBPROD
CurveType SYMBPROD( CurveType Crv1, CurveType Crv2 )
or
SurfaceType SYMBPROD( SurfaceType Srf1, SurfaceType Srf2 )
Computes the symbolic product of the given two curves or surfaces as
a curve or surface. The product is computed coordinate-wise.
Example:
ProdSrf = SYMBPROD( Srf1, Srf2 )
$
SYMBDPROD
CurveType SYMBDPROD( CurveType Crv1, CurveType Crv2 )
or
SurfaceType SYMBDPROD( SurfaceType Srf1, SurfaceType Srf2 )
Computes the symbolic dot (inner) product of the given two curves or surfaces
as a scalar curve or surface.
Example:
DiffCrv = symbdiff( Crv1, Crv2 )
DistSqrCrv = SYMBDPROD( DiffCrv, DiffCrv )
Computes a scalar curve that at parameter t is equal to the distance
square between Crv1 at t and Crv2.
$
SYMBCPROD
CurveType SYMBCPROD( CurveType Crv1, CurveType Crv2 )
or
SurfaceType SYMBCPROD( SurfaceType Srf1, SurfaceType Srf2 )
Computes the symbolic cross product of the given two curves or surfaces as
a curve or surface.
Example:
NrmlSrf = SYMBCPROD( sderive( Srf, ROW ), sderive( Srf, COL ) )
computes a normal surface as the cross product of the surface two partial
derivatives (see SNRMLSRF).
$
SYMBSUM
CurveType SYMBSUM( CurveType Crv1, CurveType Crv2 )
or
SurfaceType SYMBSUM( SurfaceType Srf1, SurfaceType Srf2 )
Computes the symbolic sum of the given two curves or surfaces as
a curve or surface. The sum is computed coordinate-wise.
Example:
SumCrv = SYMBSUM( Crv1, Crv2 )
$
SYMBDIFF
CurveType SYMBDIFF( CurveType Crv1, CurveType Crv2 )
or
SurfaceType SYMBDIFF( SurfaceType Srf1, SurfaceType Srf2 )
Computes the symbolic difference of the given two curves or surfaces as
a curve or surface. The difference is computed coordinate-wise.
Example:
DiffCrv = SYMBDIFF( Crv1, Crv2 )
DistSqrCrv = symbdprod( DiffCrv, DiffCrv )
$
TBEZIER
TrivarType TBEZIER( ListType CtlMesh )
Creates a Bezier trivariate using the provided control mesh. CtlMesh
is a list of planes, each of which is a list of rows, each of which is a list
of control points. All control points must be of type (E1-E5, P1-P5), or
regular PointType defining the trivariate's control mesh. Surface's point
type will be of a space which is the union of the spaces of all points.
Example:
TV = TBEZIER( list( list( list( ctlpt( E3, 0.1, 0.1, 0.0 ),
ctlpt( E3, 0.2, 0.5, 1.1 ),
ctlpt( E3, 0.3, 0.1, 2.2 ) ),
list( ctlpt( E3, 0.4, 1.3, 0.5 ),
ctlpt( E3, 0.5, 1.7, 1.7 ),
ctlpt( E3, 0.6, 1.3, 2.9 ) ),
list( ctlpt( E3, 0.7, 2.4, 0.5 ),
ctlpt( E3, 0.8, 2.6, 1.4 ),
ctlpt( E3, 0.9, 2.8, 2.3 ) ) ),
list( list( ctlpt( E3, 1.1, 0.1, 0.5 ),
ctlpt( E3, 1.3, 0.2, 1.7 ),
ctlpt( E3, 1.5, 0.3, 2.9 ) ),
list( ctlpt( E3, 1.7, 1.2, 0.0 ),
ctlpt( E3, 1.9, 1.4, 1.2 ),
ctlpt( E3, 1.2, 1.6, 2.4 ) ),
list( ctlpt( E3, 1.4, 2.3, 0.9 ),
ctlpt( E3, 1.6, 2.5, 1.7 ),
ctlpt( E3, 1.8, 2.7, 2.5 ) ) ) ) );
creats a trivariate Bezier which is linear in the first direction, and
quadratic in the second and third.
$
TBSPLINE
TrivarType TBSPLINE( NumericType UOrder,
NumericType VOrder,
NumericType WOrder,
ListType CtlMesh,
ListType KnotVectors )
Creates a Bspline trivariate with the provided UOrder, VOrder
and WOrder orders, the control mesh CtlMesh, and the three knot
vectors in KnotVectors.
CtlMesh is a list of planes, each of which is a list of rows, each
of which is a list of control points. All control points must be of point
type (E1-E5, P1-P5), or regular PointType defining the trivariate's control
mesh. Trivariate's point type will be of a space which is the union of the
spaces of all points.
KnotVectors is a list of three knot vectors. Each knot vector is a
list of NumericType knots of length #CtlPtList plus the Order.
If, however, the length of the knot vector is equal to #CtlPtList +
Order + Order - 1} the curve is assumed periodic.
The knot vector may also be a list of a single constant KV_OPEN or
KV_FLOAT or KV_PERIODIC, in which a uniform knot vector with the
appropriate length and with open, floating or periodic end condition
will be constructed automatically.
Example:
TV = TBSPLINE( 2, 2, 2,
list( list( list( ctlpt( E3, 0.1, 0.1, 0.0 ),
ctlpt( E3, 0.2, 0.5, 1.1 ),
ctlpt( E3, 0.3, 0.1, 2.2 ) ),
list( ctlpt( E3, 0.4, 1.3, 0.5 ),
ctlpt( E3, 0.5, 1.7, 1.7 ),
ctlpt( E3, 0.6, 1.3, 2.9 ) ),
list( ctlpt( E3, 0.7, 2.4, 0.5 ),
ctlpt( E3, 0.8, 2.6, 1.4 ),
ctlpt( E3, 0.9, 2.8, 2.3 ) ) ),
list( list( ctlpt( E3, 1.1, 0.1, 0.5 ),
ctlpt( E3, 1.3, 0.2, 1.7 ),
ctlpt( E3, 1.5, 0.3, 2.9 ) ),
list( ctlpt( E3, 1.7, 1.2, 0.0 ),
ctlpt( E3, 1.9, 1.4, 1.2 ),
ctlpt( E3, 1.2, 1.6, 2.4 ) ),
list( ctlpt( E3, 1.4, 2.3, 0.9 ),
ctlpt( E3, 1.6, 2.5, 1.7 ),
ctlpt( E3, 1.8, 2.7, 2.5 ) ) ) ),
list( list( KV_OPEN ),
list( KV_OPEN ),
list( KV_OPEN ) ) );
constructs a trilinear Bspline trivariate with open end conditions.
$
TCRVTR
TCRVTR
AnyType TCRVTR( TrivarType TV, PointType Pos, NumericType ComputeWhat )
Compute differential curvature properties of an isosurface of
the given trivariate TV at the given (parameteric) location
Pos. Following the value of ComputeWhat, the result equals,
-1 Initialization (a must prelude)
0 Conclusion (a must postlude)
1 Returns a vector hold the gradient
2 Returns a list of three vectors
equal to the Hessian of this location
3 Returns a list of two scalar values
(Principal curvatures) and two vectors
(Principal directions).
Every evaluation must start with an invokation with ComputeWhat
equal -1 and terminates with ComputeWhat 0. In both cases, 1 is
return in case of success.
Example:
TCRVTR( TV, point( 0, 0, 0 ), -1 ); # Prelude
Grad1 = TCRVTR( TV, point( 0, 0, 0 ), 1 );
Grad2 = TCRVTR( TV, point( 0, 0, 1 ), 1 );
Grad3 = TCRVTR( TV, point( 0, 1, 0 ), 1 );
Grad4 = TCRVTR( TV, point( 1, 0, 0 ), 1 );
TCRVTR( TV, point( 0, 0, 0 ), 0 ); #Postlude
$
TDERIVE
TrivarType TDERIVE( TrivarType TV, NumericType Dir )
Returns a vector field trivariate representing the differentiated
trivariate in the given direction (ROW, COL, or DEPTH). Evaluation of the
returned trivariate at a given parameter value will return a vector
representing the partial derivative of TV in Dir at that
parameter value.
TV = tbezier( list( list( list( ctlpt( E1, 0.1 ),
ctlpt( E1, 0.2 ) ),
list( ctlpt( E1, 0.3 ),
ctlpt( E1, 0.4 ) ) ),
list( list( ctlpt( E1, 2.4 ),
ctlpt( E1, 2.2 ) ),
list( ctlpt( E1, 2.3 ),
ctlpt( E1, 2.1 ) ) ) ) );
DuTV = TDERIVE( TV, ROW );
DvTV = TDERIVE( TV, COL );
DwTV = TDERIVE( TV, DEPTH );
computes the gradiate of a scalar trivariate field, by computing its
partials with respect to u, v, and w.
$
TDIVIDE
TrivarType TDIVIDE( TrivarType TV, ConstantType Direction,
NumericType Param )
Subdivides a trivariate into two at the specified parameter value Param
in the specified Direction (ROW, COL, or DEPTH). TV can be either
a Bspline trivairate in which Param must be conatined in the parametric
domain of the trivariate, or a Bezier trivariate in which Param must be
in the range of zero to one.
It returns a list of the two sub-trivariates. The individual trivariates
may be extracted from the list using the NTH command.
Example:
TvDiv = TDIVIDE( Tv2, depth, 0.3 );
Tv2a = nth( TvDiv, 1 ) * tx( -2.2 );
Tv2b = nth( TvDiv, 2 ) * tx( 2.0 );
subdivides Tv2 at the parameter value of 0.3 in the DEPTH direction,
$
TEDITPT
TrivarType TEDITPT( TrivarType TV, CtlPtType CPt, NumericType UIndex,
NumericType VIndex )
NumericType WIndex )
Provides a simple mechanism to manually modify a single control point number
UIndex, VIndex and WIndex (base count is 0) in the control mesh
of Srf by substituting CtlPt instead. CtlPt must have the same
point type as the control points of Srf. Original surface Srf is
not modified.
Example:
CPt = ctlpt( E3, 1, 2, 3 );
NewTV = TEDITPT( TV, CPt, 0, 0, 0 );
Constructs a NewTV with the first control point of TV being
CPt.
$
TEVAL
CtlPtType TEVAL( TrivarType TV,
NumericType UParam,
NumericType VParam,
NumericType WParam )
Evaluates the provided trivariate TV at the given UParam,
VParam and WParam values. UParam, VParam,
WParam must be contained in the surface parametric domain if TV
is a Bspline surface, or between zero and one if TV is a Bezier
trivariate. The returned control point has the same type as the control
points of TV.
Example:
CPt = TEVAL( TV1, 0.25, 0.22, 0.7 );
evaluates TV at the parameter values of (0.25, 0.22, 0.7).
$
TEXTGEOM
AnyType TEXTGEOM( StringType Str, VectorType Spacing, NumericType Scaling )
Creates a displayable geometry that represents the text in Str, with
Spacing space between individual characters. Each character is scaled
by Scaling where scaling of one generates a close to a unit size
character.
Example:
a = TEXTGEOM("Text", vector( 0.12, 0, 0 ), 0.1 );
b = TEXTGEOM("IRIT", vector( 0, -0.12, 0 ), 0.1 );
Creates an horizontal Text and a vertical top to bottom IRIT,
both as geometrical objects.
$
TFROMSRFS
TrivarType TFROMSRFS( ListType SrfList, NumericType OtherOrder )
Constructs a trivariate by substituting the surfaces in SrfList as
planes in a control mesh of a trivariate. Surfaces in SrfList are
made compatible by promoting Bezier surfaces to Bsplines if necessary,
and raising degree and refining as required before substituting the
control meshes of the surfaces as planes in the mesh of the trivariate.
The other, third, direction order is set by OtherOrder, which cannot
be larger than the number of surfaces.
The trivariate interpolates the first and last surfaces only.
Example:
s1 = sbezier( list( list( ctlpt( E3, -0.5, -0.5, 0 ),
ctlpt( E3, -0.5, 0.5, 0 ) ),
list( ctlpt( E3, 0.5, -0.5, 0 ),
ctlpt( E3, 0.5, 0.5, 0 ) ) ) ) * sc( 0.3 );
Srfs = list( s1 * sc( 2.0 ),
s1 * sx( 1.4 ) * ry( 45 ) * tz( 1.0 ),
s1 * ry( 90 ) * trans( vector( 1.0, 0.0, 1.1 ) ),
s1 * sx( 1.4 ) * ry( 135 ) * trans( vector( 2.0, 0.0, 1.0 ) ),
s1 * sc( 2.0 ) * ry( 180 ) * trans( vector( 2.0, 0.0, 0.0 ) ) );
color( Srfs, red );
ts = tfromsrfs( Srfs, 3 );
color( ts, green );
view( list( Srfs, ts ), on );
Constructs a trivariate from five planar surfaces and display both the
trivariate and the five planar surfaces, in different colors.
$
TINTERP
TrivarType TINTERP( TrivarType TV );
Given a trivariate data structure, computes a new trivariate in the
same function space (i.e. same knot sequences and orders) that interpolates
the given triavriate, TV, at the node parameter values.
Example:
tv = tbspline( 3, 3, 2,
list( list( list( ctlpt( E3, 0.1, 0.1, 0.0 ),
ctlpt( E3, 0.2, 0.5, 1.1 ),
ctlpt( E3, 0.3, 0.1, 2.2 ) ),
list( ctlpt( E3, 0.4, 1.3, 0.5 ),
ctlpt( E3, 0.5, 1.7, 1.7 ),
ctlpt( E3, 0.6, 1.3, 2.9 ) ),
list( ctlpt( E3, 0.7, 2.4, 0.5 ),
ctlpt( E3, 0.8, 2.6, 1.4 ),
ctlpt( E3, 0.9, 2.8, 2.3 ) ) ),
list( list( ctlpt( E3, 1.1, 0.1, 0.5 ),
ctlpt( E3, 1.3, 0.2, 1.7 ),
ctlpt( E3, 1.5, 0.3, 2.9 ) ),
list( ctlpt( E3, 1.7, 1.2, 0.0 ),
ctlpt( E3, 1.9, 1.4, 1.2 ),
ctlpt( E3, 1.2, 1.6, 2.4 ) ),
list( ctlpt( E3, 1.4, 2.3, 0.9 ),
ctlpt( E3, 1.6, 2.5, 1.7 ),
ctlpt( E3, 1.8, 2.7, 2.5 ) ) ),
list( list( ctlpt( E3, 2.8, 0.1, 0.4 ),
ctlpt( E3, 2.6, 0.7, 1.3 ),
ctlpt( E3, 2.4, 0.2, 2.2 ) ),
list( ctlpt( E3, 2.2, 1.1, 0.4 ),
ctlpt( E3, 2.9, 1.2, 1.5 ),
ctlpt( E3, 2.7, 1.3, 2.6 ) ),
list( ctlpt( E3, 2.5, 2.9, 0.7 ),
ctlpt( E3, 2.3, 2.8, 1.7 ),
ctlpt( E3, 2.1, 2.7, 2.7 ) ) ) ),
list( list( KV_OPEN ),
list( KV_OPEN ),
list( KV_OPEN ) ) );
tvi = TINTERP( tv );
creates a quadratic by quaratic by linear trivairatiate tvi that
interpolates the control points of tv at the node parameter values.
$
TMORPH
TrivarType TMORPH( TrivarType TV1, SurfaceType TV2, NumericType Blend )
Creates a new trivariate which is a convex blend of the two given
trivariates. The two given trivariates must be compatible (see FFCOMPAT)
before this blend is invoked. Very useful if a sequence that "morphs" one
trivariate to another is to be created and in combination of MRCHCUBE.
Example:
Size = 0.05;
for ( i = 0, step, 1.0,
Tv = TMORPH( Tv1, Tv2, i ):
view( mrchcube( list( Tv, 1, off ),
point( Size, Size, Size ), 1, IsoVal ), on ) );
creates a sequence of 1/step trivariates, morphed from Tv1 to Tv2
and displays an extracted iso surface at level IsoVal.
See also MRCHCUBE, CMORPH and SMORPH.
$
TORUS
PolygonType TORUS( VectorType Center, VectorType Normal,
NumericType MRadius, NumericType mRadius )
Creates a TORUS geometric object, defined by Center as the center
of the TORUS, Normal as the normal to the main plane of the TORUS,
MRadius and mRadius as the major and minor radii of the TORUS.
See RESOLUTION for the accuracy of the TORUS approximation as a polygonal
model. See IRITSTATE's "PrimRatSrfs" and "PrimRatSrfs" state variables.
Example:
T = TORUS( vector( 0.0, 0.0, 0.0), vector( 0.0, 0.0, 1.0), 0.5, 0.2 );
constructs a torus with major plane as the XY plane, major radius of 0.5,
and minor radius of 0.2.
$
TRAISE
TrivarType TRAISE( TrivarType TV, ConstantType Direction,
NumericType NewOrder )
Raise TV to the specified NewOrder in the specified
Direction.
Example:
tv1r = TRAISE( traise( traise( tv1, row, 4 ), col, 4 ), depth, 4 );
Ensures that the trivariate is a tricubic.
$
TREFINE
TrivarType TREFINE( TrivarType TV, ConstantType Direction,
NumericType Replace, ListType KnotList )
Provides the ability to Replace a knot vector of TV or refine
it in the specified direction Direction (ROW, COL, or DEPTH).
KnotList is a list of knots to refine TV at. All knots should be
contained in the parametric domain of TV in Direction. If the knot
vector is replaced, the length of KnotList should be identical to the
length of the original knot vector of TV in Direction. If TV
is a Bezier trivariate, it is automatically promoted to be a Bspline
trivariate.
Example:
TV = TREFINE( TREFINE( TREFINE( TV,
ROW, FALSE, list( 0.333, 0.667 ) ),
COL, FALSE, list( 0.333, 0.667 ) ),
DEPTH, FALSE, list( 0.333, 0.667 ) );
refines TV in all directions by adding two more knots at 0.333 and
0.667.
$
TREGION
TrivarType TREGION( TrivarType TV, ConstantType Direction,
NumericType MinParam, NumericType MaxParam )
Extracts a region of TV between MinParam and MaxParam
in the specified Direction. Both MinParam and MaxParam
should be contained in the parametric domain of TV in Direction.
Example:
Tv1 = tbezier( list( list( list( ctlpt( E3, 0.1, 0.0, 0.8 ),
ctlpt( E3, 0.2, 0.1, 2.4 ) ),
list( ctlpt( E3, 0.3, 2.2, 0.2 ),
ctlpt( E3, 0.4, 2.3, 2.0 ) ) ),
list( list( ctlpt( E3, 2.4, 0.8, 0.1 ),
ctlpt( E3, 2.2, 0.7, 2.3 ) ),
list( ctlpt( E3, 2.3, 2.6, 0.5 ),
ctlpt( E3, 2.1, 2.5, 2.7) ) ) ) );
Tv1r1 = TREGION( Tv1, row, 0.1, 0.2 );
Tv1r2 = TREGION( Tv1, row, 0.4, 0.6 );
Tv1r3 = TREGION( Tv1, row, 0.99, 1.0 );
extracts three regions of Tv1 along the ROW direction.
$
TREPARAM
TrivarType TREPARAM( TrivarType TV, ConstantType Direction,
NumericType MinParam, NumericType MaxParam )
Reparametrize TV over a new domain from MinParam to MaxParam,
in the prescribed Direction. This operation does not affect the geometry
of the trivariate and only affine transforms its knot vectors. A Bezier
trivariate will automatically be promoted into a Bspline surface by this
function.
Example:
Tv = TREPARAM( TREPARAM( TREPARAM( tv, row, 0, 1 ),
col, 0, 1 ),
depth, 0, 1 );
Ensures that the trivariate is defined over the unit size parametric
cube.
$
TRIANGL
PolygonType TRIANGL( PolygonType Model )
Converts Model into a new model with exactly the same shape that holds
only triangles.
$
TRIMSRF
TrimSrfType TRIMSRF( SurfaceType Srf,
CurveType TrimCrv,
NumericType HasUpperLevel )
or
TrimSrfType TRIMSRF( SurfaceType Srf,
ListType TrimCrvs,
NumericType HasUpperLevel )
Creates a trimmed surface from the provided surface Srf and the
trimming curve TrimCrv or curves TrimCrvs. If HasUpperLevel
is FALSE, an additional trimming curve is automatically being added that
contains the entire parametric domain of Srf. No validity test is
performed on the trimming curves which are assumed two dimensional curves
contained in the parametric domain of Srf.
Example:
spts = list( list( ctlpt( E3, 0.1, 0.0, 1.0 ),
ctlpt( E3, 0.3, 1.0, 0.0 ),
ctlpt( E3, 0.0, 2.0, 1.0 ) ),
list( ctlpt( E3, 1.1, 0.0, 0.0 ),
ctlpt( E3, 1.3, 1.5, 2.0 ),
ctlpt( E3, 1.0, 2.1, 0.0 ) ),
list( ctlpt( E3, 2.1, 0.0, 2.0 ),
ctlpt( E3, 2.3, 1.0, 0.0 ),
ctlpt( E3, 2.0, 2.0, 2.0 ) ),
list( ctlpt( E3, 3.1, 0.0, 0.0 ),
ctlpt( E3, 3.3, 1.5, 2.0 ),
ctlpt( E3, 3.0, 2.1, 0.0 ) ),
list( ctlpt( E3, 4.1, 0.0, 1.0 ),
ctlpt( E3, 4.3, 1.0, 0.0 ),
ctlpt( E3, 4.0, 2.0, 1.0 ) ) );
sb = sbspline( 3, 3, spts, list( list( KV_OPEN ), list( KV_OPEN ) ) );
TCrv1 = cbspline( 2,
list( ctlpt( E2, 0.3, 0.3 ),
ctlpt( E2, 0.7, 0.3 ),
ctlpt( E2, 0.7, 0.7 ),
ctlpt( E2, 0.3, 0.7 ),
ctlpt( E2, 0.3, 0.3 ) ),
list( KV_OPEN ) );
TCrv2 = circle( vector( 0.5, 0.5, 0.0 ), 0.25 );
TCrv3 = cbspline( 3,
list( ctlpt( E2, 0.3, 0.3 ),
ctlpt( E2, 0.7, 0.3 ),
ctlpt( E2, 0.7, 0.7 ),
ctlpt( E2, 0.3, 0.7 ) ),
list( KV_PERIODIC ) );
TSrf1 = TRIMSRF( sb, TCrv1, false );
TSrf2 = TRIMSRF( sb, TCrv1, true );
TSrf3 = TRIMSRF( sb, list( TCrv1, TcRv2 * ty( 1 ), TCrv3 * ty( 2 ) ),
false );
constructs three trimmed surfaces. Tsrf1 contains the outer boundary
and excludes what is inside TCrv1, TSrf2 contains only the
domain inside TCrv1. TCrv3 has three holes corresponds to the
three trimming curves.
$
TSBEZIER
SurfaceType TSBEZIER( NumericType Order, ListType CtlMesh )
Creates a triangular Bezier surface of oder Order using the
provided control mesh. CtlMesh is a list of control points of
size (Order + 1) * Order / 2. All control points
must be of type (E1-E5, P1-P5), or regular PointType defining the surface's
control polygon. Surface's point type will be of a space which is the union
of the spaces of all points.
Example:
b = TSBEZIER( 3,
list( ctlpt( E3, 0.0, 0.0, 0.4 ),
ctlpt( E3, 0.3, 0.0, 0.3 ),
ctlpt( E3, 0.7, 0.0, 0.8 ),
ctlpt( E3, 0.2, 0.4, 1.0 ),
ctlpt( E3, 0.4, 0.5, 1.0 ),
ctlpt( E3, 0.5, 1.0, 0.7 ) ) );
$
TSBSPLINE
TriSrfType TSBSPLINE( NumericType Order, NumericType Length,
ListType CtlMesh, ListType KnotVector )
Creates a Bspline surface from the provided Order and Length,
the control mesh CtlMesh, and the knot vector KnotVector.
CtlMesh is a list of control points of size (Length + 1) *
Length / 2. All control points must be of point type (E1-E5,
P1-P5), or regular PointType defining the surface's control mesh.
Surface's point type will be of a space which is the union of the spaces
of all points.
KnotVector is a list of NumericType knots of length Length
plus the Order.
The knot vector may also be a list of a single constant KV_OPEN or
KV_FLOAT, in which a uniform knot vector with the appropriate length
and with open or floating end condition will be constructed automatically.
Not fully supported at this time.
$
TSDERIVE
TriSrfType TSDERIVE( TriSrfType Srf, NumericType Dir )
Returns a vector field surface representing the differentiated triangular
surface in the given direction (ROW, COL, or DEPTH). Evaluation of
the returned surface at a given parameter value will return a vector
tangent to Srf in Dir at that parameter value.
DuSrf = TSDERIVE( Srf, ROW );
DvSrf = TSDERIVE( Srf, COL );
Normal = coerce( tseval( DuSrf, 0.5, 0.25, 0.25 ), VECTOR_TYPE ) ^
coerce( tseval( DvSrf, 0.5, 0.25, 0.25 ), VECTOR_TYPE );
computes two partial derivatives of the surface Srf and computes
its normal as their cross product, at the parametric location
(0.5, 0.25, 0.25).
See also TSNORMAL
$
TSEVAL
CtlPtType TSEVAL( TriSrfType Srf,
NumericType UParam, NumericType VParam, NumericType WParam )
Evaluates the provided triangulae surface Srf at the given
UParam, VParam, WParam parameters.
UParam, VParam, and WParam must all be non negative
and must sum to one for a Bezier triangular surface or to the maximum
domain, if Bspline surface.
Example:
CPt = TSEVAL( Srf, u, v, 1.0 - u - v );
Evaluates Srf at the parameter values prescribed by u and v.
$
TSNORMAL
VectorType TSNORMAL( TriSrfType Srf,
NumericType UParam, NumericType VParam, NumericType WParam)
Computes the normal vector to a triangular surface Srf at the
parameter values UParam, VParam, WParam. The returned
vector has a unit length.
UParam, VParam, and WParam must all be non
negative and must sum to one for a Bezier triangular surface or to
the maximum domain, if Bspline surface.
Example:
Normal = TSNORMAL( Srf, 0.5, 0.5, 0.0 );
computes the normal to Srf at the parameter values (0.5, 0.5, 0.0).
$
TVLOAD
TrivarType TVLOAD( StringType FileName,
NumericType DataType,
VectorType VolSize,
VectorType Orders )
Loads a volumetric data set from file FileName in as a
trivariate of orders Orders. DataType can be one of:
1 Regular ascii (separated by while spaces).
2 Two bytes short integer.
3 Four bytes long integer.
4 One byte (char) integer.
5 Four bytes float.
6 Eight bytes double.
Beware of the little vs big Endian problem! We assume here you read in the
volume in the same machine type this file was writeen with.
VolSize provides the dimensions of the volume, with width first and
depth last. Uniform open end condition knot vector are constructed to all
three axes.
Example:
Tv = tvload( "3dhead", 1, vector( 32, 32, 13 ), vector( 3, 3, 3 ) );
loads the data set "3dhead" of size (32, 32, 13) as a triquadratic
function. Data set is assumed to contain ascii numeric values.
See also MRCHCUBE.
$
Object transformation functions
All the routines in this section construct a 4 by 4 homogeneous
transformation matrix representing the required transform. These matrices
may be concatenated to achieve more complex transforms using the matrix
multiplication operator *. For example, the expression
m = trans( vector( -1, 0, 0 ) ) * rotx( 45 ) * trans( vector( 1, 0, 0 ) );
constructs a transform to rotate an object around the X = 1 line, 45 degrees.
A matrix representing the inverse transformation can be computed as:
InvM = m ^ -1
See also overloading of the - operator.
$
HOMOMAT
MatrixType HOMOMAT( ListType MatData )
Creates an arbitrary homogeneous transformation matrix by manually providing
its 16 coefficients.
Example:
for ( a = 1, 1, 720 / step,
view_mat = save_mat *
HOMOMAT( list( list( 1, 0, 0, 0 ),
list( 0, 1, 0, 0 ),
list( 0, 0, 1, -a * step / 500 ),
list( 0, 0, 0, 1 ) ) ):
view( list( view_mat, axes ), on )
);
looping and viewing through a sequence of perspective transforms, created
using the HOMOMAT constructor.
$
ROTVEC
MatrixType ROTVEC( VectorType Vec, NumericType Angle )
Creates a rotation around the vector Vec matrix with Angle degrees.
$
ROTX
MatrixType ROTX( NumericType Angle )
Creates a rotation around the X transformation matrix with Angle degrees.
$
ROTY
MatrixType ROTY( NumericType Angle )
Creates a rotation around te Y transformation matrix with Angle degrees.
$
ROTZ
MatrixType ROTZ( NumericType Angle )
Creates a rotation around the Z transformation matrix with Angle degrees.
$
ROTZ2V
MatrixType ROTZ2V( VectorType Dir )
Creates a rotation matrix that takes Z axis into Dir. Length of Dir
is ignored.
$
ROTZ2V2
MatrixType ROTZ2V2( VectorType Dir, VectorType Dir2 )
Creates a rotation matrix that takes Z axis into Dir, while the X axis
is aligned with Dir2. The lengths of Dir and Dir2 are ignored.
$
SCALE
MatrixType SCALE( VectorType ScaleFactors )
Creates a scaling by the ScaleFactors transformation matrix.
$
TRANS
MatrixType TRANS( VectorType TransFactors )
Creates a translation by the TransFactors transformation matrix.
$
General purpose functions
$
ATTRIB
ATTRIB( AnyType Object, StringType Name, AnyType Value )
Provides a mechanism to add an attribute of any type to an Object,
with name Name and value Value. This ATTRIB function is tuned
and optimized toward numeric values or strings as Value although any
other object type can be saved as attribue.
These attributes may be used to pass information to other programs about
this object, and are saved with the objects in data files.
Attributes placed on a list object or even a whole hierarchy of
objects will be propagated into all items in the list or hierarchy.
The exception to this propagation are few. The "animation" attribute that
is not propagated and is kept in the internal nodes, forming an hierachy of
animation commands for all the objects contained in the list/hierarchy.
The "invisible" attribute is saved at all levels of the hierarchy, used to
denote a complete sub tree that is invisible (yet can serve as a source
instances can point at).
For example,
ATTRIB(Glass, "rgb", "255,0,0");
ATTRIB(Glass, "refract", "1.4");
.
.
.
RMATTR(Glass, "rgb"); # Removes "rgb" attribute.
sets the RGB color and refraction index of the Glass object and
later removes the RGB attribute.
Attribute names are case insensitive. Spaces are allowed in the Value
string, as well as the double quote itself, although the latter must be
escaped:
ATTRIB(Glass, "text", "Say "this is me"");
See also RMATTR for removal of attributes as well as AWIDTH, ADWIDTH, and
COLOR.
$
ADWIDTH
ADWIDTH( GeometricType Object, NumericType DWidth )
Sets the width of the object. This display width is used in pixels in
display devices for width of line drawing, if supported by the display device.
See also ATTRIB, COLOR, and AWIDTH.
This function is equivament to using,
ATTRIB( Object, "dwidth", Width );
$
AWIDTH
AWIDTH( GeometricType Object, NumericType Width )
Sets the width of the object to one of those specified below. This width
is used in real object side dimensions in tools such as scan converters and
rendering tools for rendering lines and curves, as well as postscript.
See also ATTRIB, COLOR, and ADWIDTH.
This function is equivament to using,
ATTRIB( Object, "width", Width );
$
CHDIR
CHDIR( StringType NewDir )
Sets the current working directory to be NewDir.
$
CLNTCLOSE
CLNTCLOSE( NumericType Handler, NumericType Kill )
Closes a communication channel to a client. Handler contains the
index of the communication channel opened via CLNTEXEC. If Kill, the
client is send an exit request for it to die. Otherwise, the communication
is closed and the client is running stand alone. See also CLNTREAD,
CLNTWRITE, and CLNTEXEC.
Example:
h2 = clntexec( "nuldrvs -s-" );
.
.
.
CLNTCLOSE( h2,TRUE );
closes the connection to the nuldrvs client, opened via CLNTEXEC.
$
CLNTWRITE
CLNTWRITE( NumericType Handler, AnyType Object )
Writes one object Object to a communication channel of a client.
Handler contains the index of the communication channel opened via
CLNTEXEC. See also CLNTREAD, CLNTCLOSE, and CLNTEXEC.
Example:
h2 = clntexec( "nuldrvs -s-" );
.
.
CLNTWRITE( h2, Model );
.
.
clntclose( h2,TRUE );
writes the object named Model to client through communication channel h2.
$
COLOR
COLOR( GeometricType Object, NumericType Color )
Sets the color of the object to one of those specified below. Note that an
object has a default color (see IRIT.CFG file) according to its origin -
loaded with the LOAD command, PRIMITIVE, or BOOLEAN operation result.
The system internally supports colors (although you may have a B&W system)
and the colors recognized are:
BLACK, BLUE, GREEN, CYAN, RED, MAGENTA, YELLOW, and WHITE.
See the ATTRIB command for more fine control of colors using the RGB
attribute. See also AWIDTH and AWIDTH.
This function is equivament to using,
ATTRIB( Object, "color", Width );
$
COMMENT
COMMENT
Two types of comments are allowed:
1. One-line comment: starts anywhere in a line at the '#' character, up to
the end of the line.
2. Block comment: starts at the COMMENT keyword followed by a unique
character (anything but white space), up to the second occurrence of that
character. This is a fast way to comment out large blocks.
Example:
COMMENT
This is a comment
$
ERROR
ERROR( StringType Message);
Breaks the execution and returns to IRIT main loop, after printing
Message to the screen. May be useful in user defined function to
break execution in cases of fatal errors.
$
EXIT
EXIT();
Exits from the solid modeler. NO warning is given!
$
FOR
FOR( NumericType Start, NumericType Increment, NumericType End, AnyType Body )
Executes the Body (see below), while the FOR loop conditions hold.
Start, Increment, End are evaluated first, and the loop is executed
while <= End if Increment > 0, or while >= End if Increment < 0.
If Start is of the form "Variable = Expression", then that variable is
updated on each iteration, and can be used within the body.
The body may consist of any number of regular commands, separated by
COLONs, including nesting FOR loops to an arbitrary level.
Example:
step = 10;
rotstepx = rotx(step);
FOR ( a = 1, 1, 360 / step,
view_mat = rotstepx * view_mat:
view( list( view_mat, axes ), ON )
);
Displays axes with a view direction that is rotated 10 degrees at a
time around the X axis.
$
HELP
HELP( StringType Subject )
Provides help on the specified Subject.
Example:
HELP("");
will list all IRIT help subjects.
$
FREE
FREE( GeometricType Object )
Because of the usually huge size of geometric objects, this procedure
may be used to free them. Reassigning a value (even of different type)
to a variable automatically releases the old variable's allocated space
as well.
$
FUNCTION
FuncName = FUNCTION(Prm1, Prm2, ... , PrmN):LclVal1:LclVar2: ... :LclVarM:
FuncBody;
Defines a function named FuncName with N parameters and M local variables
(N, M >= 0). Here is a (simple) example of a function with no local variables
and a single parameter that computes the square of a number:
sqr = FUNCTION(x):
return = x * x;
Functions can be defined with optional parameters and optional local
variables. A function's body may contain an arbitrary set of expressions
including for/while loops, (user) function calls, or even recursive function
calls, all separated by colons.
The returned value of the function is the value of an automatically defined
local variable named return. The return variable is a regular local variable
within the scope of the function and can be used as any other variable.
If a variable's name is found in neither the local variable list nor
the parameter list, it is searched in the global variable list (outside
the scope of the function). Binding of names of variables is static as in the
C programming language.
Because binding of variables is performed in execution time, there is a
somewhat less restrictive type checking of parameters of functions that are
invoked within a user's defined function.
A function can invoke itself, i.e., it can be recursive. However, since a
function should be defined when it is called, a dummy function should be
defined before the recursive one is defined:
factorial = function(x):return = x; # Dummy function.
factorial = function(x):
if (x <= 1, return = 1, return = x * factorial(x - 1));
Overloading is valid inside a function as it is outside. For example, for
add = FUNCTION(x, y):
return = x + y;
the following function calls are all valid:
add(1, 2);
add(vector(1,2,3), point(1,2,3));
add(box(vector(-3, -2, -1), 6, 4, 2), box(vector(-4, -3, -2), 2, 2, 4));
Finally, here is a more interesting example that computes an approximation
of the length of a curve, using the sqr function defined above:
distptpt = FUNCTION(pt1, pt2):
return = sqrt(sqr(coord(pt1, 1) - coord(pt2, 1)) +
sqr(coord(pt1, 2) - coord(pt2, 2)) +
sqr(coord(pt1, 3) - coord(pt2, 3)));
crvlength = FUNCTION(crv, n):pd:t:t1:t2:dt:pt1:pt2:i:
return = 0.0:
pd = pdomain(crv):
t1 = nth(pd, 1):
t2 = nth(pd, 2):
dt = (t2 - t1) / n:
pt1 = coerce(ceval(crv, t1), e3):
for (i = 1, 1, n,
pt2 = coerce(ceval(crv, t1 + dt * i), e3):
return = return + distptpt(pt1, pt2):
pt1 = pt2);
Try, for example:
crvlength(circle(vector(0.0, 0.0, 0.0), 1.0), 30) / 2;
crvlength(circle(vector(0.0, 0.0, 0.0), 1.0), 100) / 2;
crvlength(circle(vector(0.0, 0.0, 0.0), 1.0), 300) / 2;
See PROCEDURE and IRITSTATE's "DebugFunc" for more.
$
IF
IF( NumericType Cond, AnyType TrueBody { , AnyType FalseBody } )
Executes TrueBody (group of regular commands, separated by COLONs -
see FOR loop) if the Cond holds, i.e., it is a numeric value other than
zero, or optionally, if it exists, executes FalseBody if the Cond
does not hold, i.e., it evaluates to a numeric value equal to zero.
Examples:
IF ( machine == IBMOS2, resolution = 5, resolution = 10 );
IF ( a > b, max = a, max = b );
sets the resolution to be 10, unless running on an IBMOS2 system, in which
case the RESOLUTION variable will be set to 5 in the first statement, and
set max to the maximum of a and b in the second statement.
$
INCLUDE
INCLUDE( StringType FileName )
Executes the script file FileName. Nesting of include file is allowed up
to 10 levels deep. If an error occurs, all open files in all nested files
are closed and data are waited for at the top level (standard input).
A script file can contain any command the solid modeler supports.
Example:
INCLUDE( "general.irt" );
includes the file "general.irt".
$
IRITSTATE
IRITSTATE( StringType State, AnyType Data )
Sets a state variable in the IRIT solid modeller. Current supported
state variables are,
State Name Data Type Comments
Coplanar NumericType If TRUE, Coplanar polygons are handled by
Boolean
DebugMalloc StringType If "Reset", memory allocation is cleared/reset.
No "Free unallocated pointer" test after
"Reset". If "Print", all allocated blocks are
printed. Otherwise, used as "address, n": ptr
address to search for with abort() called after
n mallocs.
DebugFunc NumericType >0 user func. debug information. >2 print params
on entry, ret. val. on exit. >4 global var. list
operations.
DumpLevel NumericType Controls the way variables/expressions are
dumped. Only object names/types if >= 0, Scalars
and vectors are dumped if >= 1, Curves and
Surfaces are dumped if DumpLvl >= 2,
Polygons/lines are dumped if DumpLvl >= 3, and
List objects are traversed recursively if
DumpLvl >= 4.
EchoSource NumericType If TRUE, irit scripts are echoed to stdout.
FloatFrmt StringType Specifies a new printf floating point format.
InterCrv NumericType If TRUE Boolean operations creates only
intersection curves. If FALSE, full Boolean
operation results.
InterpProd ConstantType TRUE for Bspline sym. products via
interpolation FALSE for Bspline sym. products
via Bezier.
PolySort NumericType Axis of Polygon Intersection sweep in Boolean
operations: 0 for X axis, 1 for Y axis, 2 for
Z axis.
PrimRatSrfs NumericType TRUE for rational exact primitive surfaces,
FALSE for approximated polynomial (integral)
surfaces. See also PrimSrfs.
PrimSrfs NumericType TRUE for primitive construction as freeform
surfaces, FALSE for polygonal primitives.
See also PrimRatSrfs.
TrimCrvs NumericType Number of samples the higher order trimmed
curves are sampled, in piecewise linear
approximation. If zero, computed
symbolically as composition.
UVBoolean NumericType If TRUE, Boolean between surfaces returns UV
instead of Euclidean curves.
Example:
IRITSTATE( "DebugFunc", 3 );
IRITSTATE( "FloatFrmt", "%8.5lg" );
To print parameters of user defined functions on entry, and return value on
exit. Also selects a floating point printf format of "%8.5lg".
$
INTERACT
INTERACT( GeometryTreeType Object )
A user-defined function (see iritinit.irt) that does the following,
in order:
Clear the display device.
Display the given Object.
Pause for a keystroke.
This user-defined function in version 4.0 of IRIT is an
emulation of the INTERACT function that used to exist in previous versions.
Example:
INTERACT( list( view_mat, Axes, Obj ) );
displays and interacts with the object Obj and the predefined object
Axes. VIEW_MAT will be used to set the starting transformation.
See VIEW and VIEWOBJ for more.
$
LIST
ListType LIST( AnyType Elem1, AnyType Elem2, ... )
Constructs an object as a list of several other objects. Only a reference
is made to the Elements, so modifying Elem1 after being included in the list
will affect Elem1 in that list next time list is used!
Each inclusion of an object in a list increases its internal used
reference. The object is freed iff in used reference is zero.
As a result, attempt to delete a variable (using FREE) which is referenced
in a list removes the variable, but the object itself is freed only when the
list is freed.
$
LOAD
AnyType LOAD( StringType FileName )
Loads an object from the given FileName. The object may be any object
defined in the system, including lists, in which the structure is recovered
and reconstructed as well (internal objects are inserted into the global
system object list if they have names). If no file type is provided, ".dat"
is assumed.
This command can also be used to load binary files. Ascii regular data
files are usually loaded in much more time then binary files due the
the parsing required. Binary data files can be loaded directly like ascii
files in IRIT, but can only be inspected through IRIT tools such
as dat2irit. A binary data file must have a ".bdt" (Binary DaTa) type in
its name.
Under unix, compressed files can be loaded if the given file name has
a postfix of ".Z". The unix system's "zcat" will be invoked via a pipe
for that purpose.
$
LOGFILE
LOGFILE( NumericType Set )
or
LOGFILE( StringType FileName )
If Set is non zero (see TRUE/FALSE and ON/OFF), then everything
printed in the input window, will go to the log file specified in the
IRIT.CFG configuration file. This file will be created the first time
logfile is turned ON. If a string FileName is provided, it will
be used as a log file name from now on. It also closes the current log
file. A "LOGFILE( on );" must be issued after a log file name change.
Example:
LOGFILE( "Data1" );
LOGFILE( on );
printf( "Resolution = %lf\n", list( resolution ) );
LOGFILE( off );
to print the current resolution level into file Data1.
$
MSLEEP
MSLEEP( NumericType MilliSeconds )
Causes the solid modeller to sleep for the prescribed time in milliseconds.
Example:
for ( i = 1, 1, sizeof( crvs ),
c = nth( crvs, i ):
color( c, yellow ):
msleep(20):
viewobj( c )
);
Displays an animation sequence and sleeps for 20 milliseconds between
iterations.
$
NTH
AnyType NTH( ListType ListObject, NumericType Index )
Returns the Index (base count 1) element of the list ListObject.
Example:
Lst = list( a, list( b, c ), d );
Lst2 = NTH( Lst, 2 );
and now Lst2 is equal to 'list( b, c )'.
$
PAUSE
PAUSE( NumericType Flush )
Waits for a keystroke. Nice to have if a temporary stop in a middle of an
included file (see INCLUDE) is required. If Flush is TRUE, then the input
is first flushed to guarantee that the actual stop will occur.
$
PRINTF
PRINTF( StringType CtrlStr, ListType Data )
A formatted printing routine, following the concepts of the C programming
language's printf routine. CtrlStr is a string object for which
the following special '%' commands are supported:
%d, %i, %u Prints the numeric object as an integer or unsigned integer.
%o, %x, %X Prints the numeric object as an octal or hexadecimal integer.
%e, %f, %g, Prints the numeric object in several formats of
%E, %F floating point numbers.
%s Prints the string object as a string.
%pe, %pf, %pg Prints the three coordinates of the point object.
%ve, %vf, %vg Prints the three coordinates of the vector object.
%Pe, %Pf, %Pg, Prints the four coordinates of the plane object.
%De, %Df, %Dg, Prints the given object in IRIT's data file format.
All the '%' commands can include any modifier that is valid in the C
programming language printf routine, including l (long), prefix
character(s), size, etc. The point, vector, plane, and object commands
can also be modified in a similar way, to set the format of the
numeric data printed.
Also supported are the newline and tab using the backslash escape
character:
PRINTF("\\tThis is the char \"\\%\"\\n", nil());
Backslashes should be escaped themselves as can be seen in the above example.
Here are few more examples:
PRINTF("this is a string \"%s\" and this is an integer %8d.\\n",
list("STRING", 1987));
PRINTF("this is a vector [%8.5lvf]\\n", list(vector(1,2,3)));
IritState("DumpLevel", 9);
PRINTF("this is a object %8.6lDf...\\n", list(axes));
PRINTF("this is a object %10.8lDg...\\n", list(axes));
This implementation of PRINTF is somewhat different than the C programming
language's version, because the backslash always escapes the next
character during the processing stage of IRIT's parser. That is, the string
'\\tThis is the char \"\\%\"\\n'
is actually parsed by the IRIT's parser into
'\tThis is the char "\%"\n'
because this is the way the IRIT parser processes strings. The latter
string is the one that PRINTF actually see.
$
PROCEDURE
ProcName = PROCEDURE(Prm1, Prm2, ... , PrmN):LclVal1:LclVar2: ... :LclVarM:
ProcBody;
A procedure is a function that does not return a value, and therefore the
return variable (see FUNCTION) should not be used. A procedure is
identical to a function in every other way. See FUNCTION for more.
$
RMATTR
RMATTR( AnyType Object, StringType Name )
Removes attribute named Name from object Object. This function
will have no affect on Object if Object have no attribute named
Name.
See also ATTRIB.
$
SAVE
SAVE( StringType FileName, AnyType Object )
Saves the provided Object in the specified file name FileName.
No extension type is needed (ignored if specified), and ".dat" is supplied
by default. Object can be any object type, including list, in which
structure is saved recursively. See also LOAD. If a display device is
actively running at the time SAVE is invoked, its transformation matrix
will be saved with the same name but with extension type of ".mat" instead
of ".dat".
This command can also be used to save binary files. Ascii regular data
files are usually loaded in much more time then binary files due the
the parsing required. Binary data files can be loaded directly like ascii
files in IRIT, but must be inspected through IRIT tools such
as dat2irit. A binary data file must have a ".bdt" (Binary DaTa) type in
its name.
Under unix, files will be saved compressed if the given file name has
a postfix of ".Z". The unix system's "compress" will be invoked via a pipe
for that purpose.
Example:
SAVE( "Obj1.bdt.Z", Obj1 );
Saves Obj1 in the file Obj1.bdt.Z as compressed binary file.
$
SNOC
SNOC( AnyType Object, ListType ListObject )
Similar to the lisp cons operator but puts the new Object in the
end of the list ListObject instead of the beginning, in place.
Example:
Lst = list( axes );
SNOC( Srf, Lst );
and now Lst is equal to the list 'list( axes, Srf )'.
$
SYSTEM
SYSTEM( StringType Command )
Executes a system command Command. For example,
SYSTEM( "ls -l" );
$
TIME
TIME( NumericType Reset )
Returns the time in seconds from the last time TIME was called with
Reset TRUE. This time is CPU time if such support is available
from the system (times function), and is real time otherwise (time
function).
The time is automatically reset at the beginning of the execution of this
program.
Example:
Dummy = TIME( TRUE );
.
.
.
TIME( FALSE );
prints the time in seconds between the above two time function calls.
$
VARLIST
VARLIST()
List all the currently defined objects in the system.
$
VECTOR
VectorType VECTOR( NumericType X, NumericType Y, NumericType Z )
Creates a vector type object, using the three provided NumericType scalars.
$
VIEW
VIEW( GeometricTreeType Object, NumericType ClearWindow )
Displays the (geometric) object(s) as given in Object.
If ClearWindow is non-zero (see TRUE/FALSE and ON/OFF) the window is
first cleared (before drawing the objects).
Example:
VIEW( Axes, FALSE );
displays the predefined object Axes in the viewing window on top of
what is drawn already.
In version 4.0, this function is emulated (see iritinit.irt) using the
VIEWOBJ function. In order to use the current viewing matrix, VIEW_MAT
should be provided as an additional parameter. For example,
VIEW( list( view_mat, Obj ), TRUE );
However, since VIEW is a user defined function, the following will not
use VIEW_MAT as one would expect:
VIEW( view_mat, TRUE );
because VIEW_MAT will be renamed inside the VIEW user defined function to
a local (to the user defined function) variable.
In iritinit.irt one can find several other useful VIEW related functions:
VIEWCLEAR Clears all data displayed on the display device.
VIEWREMOVE Removes the object specified by name from display.
VIEWDISC Disconnects from display device (which is still running)
while allowing IRIT to connect to a new device.
VIEWEXIT Forces the display device to exit.
VIEWSAVE Request sdisplay device to save transformation matrix.
BEEP An emulation of the BEEP command of versions prior to 4.0.
VIEWSTATE Allows to change the state of the display device.
For the above VIEW related functions, only VIEWREMOVE, VIEWSAVE, and
VIEWSTATE require a parameter, which is the file name and view state
respectively. The view state can be one of several commands. See the
display device section for more.
Examples:
VIEWCLEAR();
VIEW( axes, off );
VIEWSTATE( "LngrVecs" );
VIEWSTATE( "DrawSolid" );
VIEWSAVE( "matrix1" );
VIEWREMOVE( "axes" );
VIEWDISC();
$
VIEWOBJ
VIEWOBJ( GeometricTreeType Object )
Displays the (geometric) object(s) as given in Object.
Object may be any GeometricType or a list of other
GeometricTypes nested to an arbitrary level.
Unlike IRIT versions prior to 4.0, VIEW_MAT is not explicitly used
as the transformation matrix. In order to display with a VIEW_MAT view,
VIEW_MAT should be listed as an argument (in that exact name) to
VIEWOBJ. Same is true for the perspective matrix PRSP_MAT.
Example:
VIEWOBJ( list( view_mat, Axes ) );
displays the predefined object Axes in the viewing window using the
viewing matrix VIEW_MAT.
$
WHILE
WHILE( NumericType Cond, AnyType Body )
Executes the Body (see below), while the WHILE loop conditions
Cond is evaluated into a non zero value. Cond is being
evaluated before each iteration.
The body may consist of any number of regular commands, separated by
COLONs, including nesting loops to an arbitrary level.
Example:
deg = 0;
rotstepx = rotx( 10 );
WHILE ( deg < 360,
deg = deg + 10:
view_mat = rotstepx * view_mat:
view( list( view_mat, axes ), ON )
);
Displays axes with a view direction that is rotated 10 degrees at a
time around the X axis.
$
System variables
$
System variables are predefined objects in the system. Any time IRIT is
executed, these variable are automatically defined and set to values which
are sometimes machine dependent. These are regular objects in any other
sense, including the ability to delete or overwrite them. One can modify,
delete, or introduce other objects using the IRITINIT.IRT file.
AXES
Predefined polyline object (PolylineType) that describes the XYZ axes.
$
DRAWCTLPT
Predefined Boolean variable (NumericType) that controls whether curves'
control polygons and surfaces' control meshes are drawn (TRUE) or not
(FALSE). Default is FALSE.
$
FLAT4PLY
Predefined Boolean object (NumericType) that controls the way almost flat
surface patches are converted to polygons: four polygons (TRUE) or only
two polygons (FALSE). Default value is FALSE.
$
MACHINE
Predefined numeric object (NumericType) holding the machine type as one of
the following constants: MSDOS, SGI, HP, SUN, APOLLO, UNIX, IBMOS2, IBMNT,
and AMIGA.
$
POLY_APPROX_OPT
A variable controlling the algorithm to convert surfaces to polygons.
This two digit number controls the method that is used to subdivide a
surface into polygons.
The first digit (units) can be one of:
0 Uniform sampling in a fixed grid.
1 An alternated U and V subdivision direction. Once U is
subdivided and then V is subdivided.
2 A min max subdivision direction. In other words, the
direction that minimizes the maximal error is selected.
The second digit (tenths) can be one of:
0 A fixed sized regular grid. The side of the grid is set
via the RESOLUTION variable.
1 This mode is not for general use.
2 Maximal distance between the surface and its polygonal
approximation is bounded by bilinear surface fit.
Maximal distance allowed is set via POLY_APPROX_TOL.
Recommended choice for optimal polygonization.
3 This mode is not for general use.
$
POLY_APPROX_UV
A Boolean predefined variable. If TRUE, UV values of surface polygonal
approximation are placed on attribute lists of vertices.
$
POLY_APPROX_TOL
A numeric predefined tolerance control on the distance between the surface
and its polygonal approximation in POLY_APPROX_OPT settings.
$
PRSP_MAT
Predefined matrix object (MatrixType) to hold the perspective matrix
used/set by VIEW and/or INTERACT commands. See also VIEW_MAT.
$
RESOLUTION
Predefined numeric object (NumericType) that sets the accuracy of the
polygonal primitive geometric objects and the approximation of curves and
surfaces. Holds the number of divisions a circle is divided into (with
minimum value of 4). If, for example, RESOLUTION is set to 6, then a
generated CONE will effectively be a six-sided pyramid.
Also controls the fineness of freeform curves and surfaces when they are
approximated as piecewise linear polylines, and the fineness of freeform
surfaces when they are approximated as polygons.
$
VIEW_MAT
Predefined matrix object (MatrixType) to hold the viewing matrix used/set
by VIEW and/or INTERACT commands. See also PRSP_MAT.
$
System constants
The following constants are used by the various functions of the system to
signal certain conditions. Internally, they are represented numerically,
although, in general, their exact value is unimportant and may be changed
in future versions. In the rare circumstance that you need to know their
values, simply type the constant as an expression.
Example:
MAGENTA;
$
AMIGA
A constant designating an AMIGA system, in the MACHINE variable.
$
APOLLO
A constant designating an APOLLO system, in the MACHINE variable.
$
BLACK
A constant defining a BLACK color.
$
BLUE
A constant defining a BLUE color.
$
COL
A constant defining the COLumn or U direction of a surface or a
trivariate mesh.
$
CTLPT_TYPE
A constant defining an object of type control point.
$
CURVE_TYPE
A constant defining an object of type curve.
$
CYAN
A constant defining a CYAN color.
$
DEPTH
A constant defining the DEPTH direction of a trivariate mesh.
See TBEZIER, TBSPLINE.
$
E1
A constant defining an E1 (X only coordinate) control point type.
$
E2
A constant defining an E2 (X and Y coordinates) control point type.
$
E3
A constant defining an E3 (X, Y and Z coordinates) control point type.
$
E4
A constant defining an E4 control point type.
$
E5
A constant defining an E5 control point type.
$
FALSE
A zero constant. May be used as Boolean operand.
$
GREEN
A constant defining a GREEN color.
$
HP
A constant designating an HP system, in the MACHINE variable.
$
IBMOS
A constant designating an IBM system running under OS2, in the MACHINE
variable.
$
IBMNT
A constant designating an IBM system running under Windows NT, in the MACHINE
variable.
$
KV_FLOAT
A constant defining a floating end condition uniformly spaced knot vector.
$
KV_OPEN
A constant defining an open end condition uniformly spaced knot vector.
$
KV_PERIODIC
A constant defining a periodic end condition with uniformly spaced knot
vector.
$
LIST_TYPE
A constant defining an object of type list.
$
MAGENTA
A constant defining a MAGENTA color.
$
MATRIX_TYPE
A constant defining an object of type matrix.
$
MSDOS
A constant designating an MSDOS system, in the MACHINE variable.
$
NUMERIC_TYPE
A constant defining an object of type numeric.
$
OFF
Synonym of FALSE.
$
ON
Synonym for TRUE.
$
P2
A constant defining a P1 (W and WX coordinates, in that order) rational
control point type.
$
P2
A constant defining a P2 (W, WX, and WY coordinates, in that order) rational
control point type.
$
P3
A constant defining a P3 (W, WX, WY, and WZ coordinates, in that order)
rational control point type.
$
P4
A constant defining a P4 rational control
point type.
$
P5
A constant defining a P5 rational control
point type.
$
PARAM_CENTRIP
A constant defining a centripetal length parametrization.
$
PARAM_CHORD
A constant defining a chord length parametrization.
$
PARAM_UNIFORM
A constant defining an uniform parametrization.
$
PI
The constant of 3.141592...
$
PLANE_TYPE
A constant defining an object of type plane.
$
POINT_TYPE
A constant defining an object of type point.
$
POLY_TYPE
A constant defining an object of type poly.
$
RED
A constant defining a RED color.
$
ROW
A constant defining the ROW or V direction of a surface or a trivariate mesh.
$
SGI
A constant designating an SGI system, in the MACHINE variable.
$
STRING_TYPE
A constant defining an object of type string.
$
SURFACE_TYPE
A constant defining an object of type surface.
$
SUN
A constant designating a SUN system, in the MACHINE variable.
$
TRIMSRF_TYPE
A constant defining an object of type trimmed surface.
$
TRIVAR_TYPE
A constant defining an object of type trivariate function.
$
TRUE
A non zero constant. May be used as Boolean operand.
$
UNDEF_TYPE
A constant defining an object of no type (yet).
$
UNIX
A constant designating a generic UNIX system, in the MACHINE variable.
$
VECTOR_TYPE
A constant defining an object of type vector.
$
WHITE
A constant defining a WHITE color.
$
YELLOW
A constant defining a YELLOW color.
$
Animation
The animation tool adds the capability of animating objects using
forward kinematics, exploiting animation curves. Each object has
different attributes, that prescribe its motion, scale, and visibility
as a function of time. Every attribute has a name, which designates
it's role. For instance an attribute animation curve named MOV_X
describes a translation motion along the X axis.
Let OBJ be an object in IRIT to animate.
Animation curves are either scalar (E1/P1) curves or three dimensional
(E3/P3) curves with one of the following name prefixes:
MOV_X, MOV_Y, MOV_Z Translation along one axis
MOV_XYZ Arbitrary translation along all three axes
ROT_X, ROT_Y, ROT_Z Rotating around a single axis (degrees)
SCL_X, SCL_Y, SCL_Z Scale along a single axis
SCL Global scale
VISIBLE Visibility
The visibility curve is a scalar curve that enables the display of
the object if the visibility curve is positive at time t and disables
the display (hide) the object if the visibility curve is negative at
time t.
The animation curves are all attached as an attribute named "animation"
to the object OBJ.
Example:
mov_x = cbezier( ctlpt( E1, 0.0 ),
ctlpt( E1, 1.0 ) );
scl = cbezier( ctlpt( E1, 1.0 ),
ctlpt( E1, 0.1 ) );
rot_y = cbezier( ctlpt( E1, 0.0 ),
ctlpt( E1, 0.0 ) );
ctlpt( E1, 360.0 ) );
attrib(OBJ, "animation", list( mov_x, scl, rot_y ) );
To animate OBJ between time zero and one (Bezier curves are always
between zero and one), by moving it a unit size in the X direction,
scaling it to %10 of its original size and rotating it at increasing
angular speed from zero to 360 degrees.
OBJ can now be save into a file or displayed via one of the regular
viewing commands in IRIT (i.e. VIEWOBJ).
Animation is not always between zero and one. To that end, one can
apply the CREPARAM function to modify the parametric domain of the
animation curve. The convention is that if the time is below the
starting value of the parametric domain, the starting value of the
curve is used. Similarly if the time is beyond the end of the
parameter domain of the animation curve, the end value of the
animation curve is used.
Example:
CREPARAM( mov_x, 3.0, 5.0 );
to set the time of the motion in the x axis to be from t = 3 to
t = 5. for t < 3, mov_x(3) is used, and for t > 5, mov_x(5) is
employed.
the animation curves are regular objects in the IRIT system. Hence,
only one object named mov_x or scl can exist at one time. If you
create a new object named mov_x, the old one is overwritten! To
preserve old animation curves you can detach the old ones by executing
'free(mov_x)' that removes the object named mov_x from IRIT's object
list but not from its previous used locations within other list
objects, if any. A different way to it call this animation curves
mov_x1, mov_x2 etc. as only the prefix of the name is verified.
For example:
mov_x = cbezier( ctlpt( E1, 0.0 ),
ctlpt( E1, 1.0 ) );
attrib(obj1, "animation", list( mov_x ) );
free(mov_x);
mov_x1 = cbezier( ctlpt( E1, 2.0 ),
ctlpt( E1, 3.0 ) );
mov_x2 = cbezier( ctlpt( E1, 2.0 ),
ctlpt( E1, 3.0 ) );
attrib(obj2, "animation", list( mov_x1, mov_x2 ) );
free(mov_x);
notice the way we have two animation curves translating obj2 in x.
This is somewhat artificial but might make more sense if other
transformations would appear in between.
a = box( vector( 0, 0, 0 ), 1, 1, 1 );
b = box( vector( 0, 0, 0 ), 1, 1, 1 );
c = box( vector( 0, 0, 0 ), 1, 1, 1 );
d = sphere( vector( 0, 0, 0), 0.7 );
pt0 = ctlpt( e1, 0.0 );
pt1 = ctlpt( e1, 1.0 );
pt2 = ctlpt( e1, 2.0 );
pt6 = ctlpt( e1, 6.0 );
pt360 = ctlpt( e1, 360.0 );
pt10 = ctlpt( e1, -4.0 );
pt11 = ctlpt( e1, 1.0 );
pt12 = ctlpt( e1, 4.0 );
pt13 = ctlpt( e1, -1.0 );
visible = creparam( cbezier( list( pt10, pt11 ) ), 0.0, 5.0 );
mov_x = creparam( cbezier( list( pt0, pt6, pt2 ) ), 0.0, 1.2 );
mov_y = mov_x;
mov_z = mov_x;
rot_x = creparam( cbspline( 2,
list( pt0, pt360, pt0 ),
list( KV_OPEN ) ),
1.2, 2.5 );
rot_y = rot_x;
rot_z = rot_x;
scl = creparam( cbezier( list( pt1, pt2, pt1, pt2, pt1 ) ),
2.5, 4.0 );
scl_x = scl;
scl_y = scl;
scl_z = scl;
mov_xyz = creparam( circle( vector( 0, 0, 0 ), 2.0 ), 4.0, 5.0 );
attrib( d, "animation", list( mov_xyz, visible ) );
free( visible );
visible = creparam( cbezier( list( pt12, pt13 ) ), 0.0, 5.0 );
attrib( a, "animation", list( rot_x, mov_x, scl, scl_x, visible ) );
attrib( b, "animation", list( rot_y, mov_y, scl, scl_y, visible ) );
attrib( c, "animation", list( rot_z, mov_z, scl, scl_z, visible ) );
color( a, red );
color( b, green );
color( c, blue );
color( d, cyan );
demo = list( a, b, c, d );
interact( demo );
viewanim( 0, 5, 0.01 );
In this example, we create four objects, three cubes and one sphere.
Animation curves to translate the three cubes along the three axes for
the time period of t = 0 to t = 1.2 are created. Rotation curves to
rotate the three cubes along the three axes are then created for time
period of t = 1.2 to t = 2.5. Finally, for the time period of t = 2.5
to t = 4.0. the cubes are (not only) unifomly scaled. For the time
period of t = 4 to t = 5, the cubes become invisible and the sphere,
that becomes visible, is rotated along a circle of radius 2.
This example demonstrates the ability to put "animation" attributes
on internal nodes of a hierarchy, affecting the entire set of objects
in the hierachy. Herein, we present an robotic arm with three edges
and two joints.
BoxLength = 2;
BoxWidth = 2;
BoxHeight = 10;
LowerBox = box( vector( -BoxLength / 2, -BoxWidth / 2, 0 ),
BoxLength, BoxWidth, BoxHeight);
MiddleBox = box( vector( -BoxLength / 2, -BoxWidth / 2, 0 ),
BoxLength, BoxWidth, BoxHeight);
UpperBox = box( vector( -BoxLength / 2, -BoxWidth / 2, 0 ),
BoxLength, BoxWidth, BoxHeight);
Cn1 = cone( vector( 0, 0, 0 ), vector( 0, BoxHeight / 3, 0 ), 1 );
color( LowerBox, magenta );
color( MiddleBox, yellow );
color( UpperBox, cyan );
color( Cn1, green );
rot_x1 = creparam( cbspline( 3,
list( ctlpt( E1, 0 ),
ctlpt( E1, -200 ),
ctlpt( E1, 200 ),
ctlpt( E1, 0 ) ),
list( KV_OPEN ) ),
0, 3 );
rot_x2 = creparam( cbspline( 4,
list( ctlpt( E1, 0 ),
ctlpt( E1, 400 ),
ctlpt( E1, -400 ),
ctlpt( E1, 0 ) ),
list( KV_OPEN ) ),
0, 3 );
rot_y = creparam( cbspline( 2,
list( ctlpt( E1, 0 ),
ctlpt( E1, 100 ),
ctlpt( E1, -100 ),
ctlpt( E1, 0 ) ),
list( KV_OPEN ) ),
0, 3 );
rot_z = creparam( cbspline( 2,
list( ctlpt( E1, 0 ),
ctlpt( E1, 1440 ) ),
list( KV_OPEN ) ),
0, 3 );
Translate = trans( vector( 0, 0, BoxHeight ) );
attrib( Cn1, "animation", list( rot_z, Translate ) );
Upr = list( Cn1, UpperBox );
attrib( Upr, "animation", list( rot_y, Translate ) );
Mid = list( Upr, MiddleBox );
attrib( Mid, "animation", list( rot_x2, Translate ) );
rbt_hand = list( Mid, LowerBox );
attrib( rbt_hand, "animation", list( rot_x1 ) );
view( rbt_hand, 1 );
In this example, we create four objects, three cubes and one cone,
simulating a robotic hand with three edges an a gripper (the cone).
The animation is define hierarchically, making it very easy to model
the robot.
$
Display devices
The following display device drivers are available,
Device Name Invocation Environment
xgldrvs xgldrvs -s- SGI 4D GL regular driver.
xogldrvs xogldrvs -s- SGI 4D Open GL/Motif driver.
xgladap xgladap -s- SGI 4D GL adaptive isocurve
experimental driver.
x11drvs x11drvs -s- X11 driver.
xmtdrvs xmtdrvs -s- X11 Motif driver.
xglmdrvs xglmdrvs -s- SGI 4D GL and X11/Motif driver.
wntdrvs wntdrvs -s- IBM PC Windows NT driver.
wntgdrvs wntgdrvs -s- IBM PC Windows NT Open GL driver.
os2drvs os2drvs -s- IBM PC OS2 2.x/3.x driver.
amidrvs amidrvs -s- AmigaDOS 2.04+ driver.
nuldrvs nuldrvs -s- [-d] [-D] A device to print the
object stream to stdout.
All display devices are clients communicating with the server (IRIT)
using IPC (inter process communication). On Unix and Window NT sockets are
used. A Windows NT client can talk to a server (IRIT) on a unix host if
hooked to the same netwrok. On OS2 pipes are used, and both the client and
server must run on the same machine. On AmigaDOS exec messages are used,
and both the client and server must run on the same machine.
While all display devices support object(s) transformations via a
transformation control window, many of the display devices allow one
to click and drag on the viewing window to rotate (Left Button) and
to translate (Right Button). This mode exploits the two degrees of
freedom of the mouse to provide intuitive dual axis rotation and
translation.
The server (IRIT) will automatically start a client display device
if the IRIT_DISPLAY environment variable is set to the name and options of
the display device to run. For example:
setenv IRIT_DISPLAY xgldrvs -s-
The display device must be in a directory that is in the path
environment variable. Most display devices require the '-s-' flags to run
in a non-standalone mode, or a client-server mode. Most drivers can also
be used to display data in a standalone mode (i.e., no server). For
example:
xgldrvs -s solid1.dat irit.mat
Effectively, all the display devices are also data display programs
(poly3d, which was the display program prior to version 4.0, is retired
in 4.0). Therefore some functionality is not always as expected. For
example, the quit button will always force the display device to quit,
even if poped up from IRIT, but will not cause IRIT to
quit as might logically expected. In fact, the next time IRIT will
try to communicate with the display device, it will find the broken
connection and will start up a new display device.
Most display devices recognize attributes found on objects. The following
attributes are usually recognized (depending on the device capability.):
COLOR: Selects the drawn color of the object to be one of the 8/16
predefined colors in the IRIT system: white, red, green, blue,
yellow, cyan, magenta, black.
RGB: Overwrites (if supported) the COLOR attribute (if given) and
sets the color of the object to the exact prescribed RGB set.
DWIDTH: Sets the width in pixels of the drawn object, when drawn as
a wireframe.
All display devices recognize all the command line flags and all the
configuration options in a configuration file, as described below. The display
devices will make attemps to honor the requests, to the best of their ability.
For example, only xgldrvs can render shaded models, and so only it will
honor a DrawSolid configuration options.
???drvs [-s] [-u] [-n] [-N] [-i] [-c] [-C] [-m] [-a] [-g x1,x2,y1,y2]
[-G x1,x2,y1,y2] [-I #IsoLines] [-F PolygonOpti FineNess]
[-f PolylineOpti SampTol] [-l LineWidth] [-r] [-A Shader] [-B]
[-2] [-d] [-D] [-L NormalLen] [-4] [-b BackGround] [-S LgtSrcPos]
[-Z ZMin ZMax] [-M] [-P] [-x ExecAnimCmd] [-X Min,Max,Dt{,flags}]
[-z] DFiles
-s: Runs the driver in a Standalone mode. Otherwise, the driver will
attempt to communicate with the IRIT server.
-u: Forces a Unit matrix. That is, input data are not
transformed at all.
-n: Draws normals of vertices.
-N: Draws normals of polygons.
-i: Draws internal edges (created by IRIT) - default is not to
display them, and this option will force displaying them as well.
-c: Sets depth cueing on. Drawings that are closer to the viewer will
be drawn in more intense color.
-C: Cache the piecewise linear geometry so curves and surface can
be redisplay faster. Purging it will free memory, on the other hand.
-m: Provides some more information on the data file(s) parsed.
-a: Sets the support of antialiased lines.
-g x1,x2,y1,y2: Prescribes the position and location of the
transformation window by prescribing the domain of the window in
screen space pixels.
-G x1,x2,y1,y2: Prescribes the position and location of the
viewing window by prescribing the domain of the window in
screen space pixels.
-I #IsoLines: Specifies number of isolines per surface, per direction.
A specification of zero isolines is possible only on the command line
and it denotes the obvious.
-F PolyOpti FineNess: Controls the method used to approximate surfaces
into polygons. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-l LineWidth: Sets the linewidth, in pixels. Default is one pixel wide.
-r: Rendered mode. Draws object as solid.
-A Shader: Shader can be one of 1 (Flat), 2 (Gouraud), or 3 (Phong).
-B: Back face culling of polygons.
-2: Double buffering. Prevents screen flicker on the expense of
possibly less colors.
-d: Debug objects. Prints to stderr all objects read from communcation
port with the server IRIT.
-D: Debug input. Prints to stderr all characters read from communcation
port with the server IRIT. Lowest level of communication.
-L NormalLen: Sets the length of the drawn normals in thousandths of
a unit.
-4: Forces four polygons per almost flat region in the surface to
polygon conversion. Otherwise two polygons only.
-b BackGround: Sets the background color as three RGB integers in the
range of 0 to 255.
-S LgtSrcPos: Sets the lighting via the light source position.
-Z ZMin ZMax: Sets the near and far Z clipping planes.
-M: Draw control mesh/polygon of curves and surfaces, as well.
-x ExecAnimCmd: Command to execute as a subprocess every iteration of
display of an animation sequence. This command can for example save
the display into an image file, saving the animation sequence.
One parameter is passed, which is an running index starting from one.
-X Min,Max,Dt{,flags}: Executes an animation sequence between
Min time to Max time is steps of Dt. If an 's' flag is specified
the animation is saved as idividual data files, one per frame, for
high quality rendering. If 'x' flag is given, the display device
exists once the animation is complete.
-P: Draws curves and surfaces (surfaces are drawn using a set of
isocurves, see -I, or polygons, see -f).
-z: Prints version number and current defaults.
Configuration Options
The configuration file is read before the command line options are
processed. Therefore, all options in this section can be overriden
by the appropriate command line option, if any.
TransPrefPos: Preferred location (Xmin, YMin, Xmax, Ymax) of the
transformation window.
ViewPrefPos: Preferred location (Xmin, YMin, Xmax, Ymax) of the
viewing window.
BackGround: Background color. Same as '-b'.
Internal: Draws internal edges. Same as '-i'.
LightSrcPos: Sets the location of the (first) light source as a rational
four coefficient location. W of zero sets the light source at infinity.
ExecAnimCmd: Executes a command each step of animation. Same as '-x'.
ExecAnimation: Executes an animation sequence on startup. Same as '-X'.
DrawVNormal: Draws normals of vertices. Same as '-n'.
DrawPNormal: Draws normals of polygons. Same as '-n'.
MoreVerbose: Provides some more information on the data file(s)
parsed. Same as '-m'.
UnitMatrix: Forces a Unit matrix. That is, input data are not
transformed at all. Same as '-u'.
DrawSolid: Requests a shaded surface rendering, as opposed to isocurve
surface representation.
BFaceCull: Requests the removal of back facing polygons, for better
visibility.
DoubleBuffer: Requests drawing using a double buffer, if any.
DebugObjects: Debug objects. Prints to stderr all objects read
from the communcation port with the server IRIT. Same as '-d'.
DebugEchoInput: Debug input. Prints to stderr all characters read
from the communcation port with the server IRIT. Lowest level of
communications.
DepthCue: Set depth cueing on. Drawings that are closer to the
viewer will be drawn in more intense color. Same as '-c'.
CacheGeom: Normally piecewise linear approximation of freefroms is
cached. By setting this option to FALSE, no such auxiliary data is
being saved, reducing the memory overhead. Same as '-C'.
FourPerFlat: Forces four polygons per almost flat region in the
surface to polygon conversion. Otherwise two polygons only. Same as
'-4'.
AntiAlias: Request the drawing of anti aliased lines.
DrawSurfaceMesh: Draws control mesh/polygon of curves and surfaces,
as well. Same as '-M'.
DrawSurfacePoly: Draws curves and surfaces (surfaces are drawn using
a set of isocurves, see -I, or polygons, see -f). Same as '-P'.
StandAlone: Runs the driver in a Stand alone mode. Otherwise, the
driver will attempt to communicate with the IRIT server. Same
as '-s'.
NumOfIsolines: Specifies number of isolines per surface, per
direction. Same as '-I'.
SamplesPerCurve: Specifies the samples per (iso)curve. See '-f'.
LineWidth: Sets the linewidth, in pixels. Default is one pixel
wide. Same as '-l'
AdapIsoDir: Selects the direction of the adaptive isoline
rendering.
PolygonOpti: Controls the method used to subdivide a surface into
polygons that approximate it. Same as '-F'.
PolylineOpti: Controls the method used to subdivide a curve into
polylines that approximate it. Same as '-f'.
ShadingModel: One of 1 (Flat), 2 (Gouraud), or 3 (Phong). Same as
'-A'.
TransMode: Selects between object space transformations and screen
space transformation.
ViewMode: Selects between perspective and orthographic views.
NormalLength: Sets the length of the drawn normals in thousandths of
a unit. Same as '-L'.
ZClipMin: Sets the minimal clipping plane in Z. Same as '-Z'.
ZClipMax: Sets the maximal clipping plane in Z. Same as '-Z'.
FineNess: Controls the fineness of the surface to polygon subdivision.
See '-F'.
Commands that affect the status of the display device can also be sent
via the communication port with the IRIT server. The following commands
are recognized as string objects with object name of "COMMAND_":
BEEP Makes some sound.
CLEAR Clears the display area. All objects are deleted.
DCLEAR Delayed clear. Same as CLEAR but delayed until next
object is sent from the server. Useful for animation.
DISCONNECT Closes connection with the server, but does not quit.
EXIT Closes connection with the server and quits.
GETOBJ NAME Requests the object named NAME that is returned
in the output channel to the server.
MSAVE NAME Saves the transformation matrix file by the name
NAME.
REMOVE NAME Request the removal of object named NAME from
display.
ANIMATE TMin TMax Dt Animates current scene from TMin to TMax in Dt
steps.
STATE COMMAND Changes the state of the display device. See below.
The following commands are valid for the STATE COMMAND above,
MoreSense: More sensitive mouse control.
LessSense: Less sensitive mouse control.
ScrnObject: Toggle screen/object transformation mode.
PerspOrtho: Toggles perspective/orthographic trans. mode.
DepthCue: Toggles depth cueing drawing.
DrawSolid: Toggles isocurve/shaded solid drawing.
ShadingMdl: Toggles shading model for solid solid drawing.
BFaceCull: Cull back facing polygons.
DblBuffer: Toggles single/double buffer mode.
AntiAlias: Toggles anti aliased lines.
DrawIntrnl: Toggles drawing of internal lines.
DrawVNrml: Toggles drawing of normals of vertices.
DrawPNrml: Toggles drawing of normals of polygons.
DSrfMesh: Toggles drawing of control meshes/polygons.
DSrfPoly: Toggles drawing of curves/surfaces.
4PerFlat: Toggles 2/4 polygons per flat surface regions.
MoreIso: Doubles the number of isolines in a surface.
LessIso: Halves the number of isolines in a surface.
FinrAprx: Doubles the number of samples per curve.
CrsrAprx: Halves the number of samples per curve.
LngrVecs: Doubles the length of displayed normal vectors.
ShrtrVecs: Halves the length of displayed normal vectors.
WiderLns: Doubles the width of the drawn lines.
NarrwLns: Halves the width of the drawn lines.
WiderPts: Doubles the width of the cross of drawn points.
NarrwPts: Halves the width of the cross of drawn points.
FinrAdapIso: Doubles the number of adaptive isocurves.
CrsrAdapIso: Halves the number of adaptive isocurves.
FinerRld: Doubles number of ruled surfaces in adaptive isocurves.
CrsrRld: Halves number of ruled surfaces in adaptive isocurves.
RuledSrfApx: Toggles ruled surface approx. in adaptive isocurves.
AdapIsoDir: Toggles the row/col direction of adaptive isocurves.
Front: Selects a front view.
Side: Selects a side view.
Top: Selects a top view.
Isometry: Selects an isometric view.
Clear: Clears the viewing area.
Obviously not all state options are valid for all drivers.
The IRIT server defines in iritinit.irt several user-defined
functions that exercise some of the above state commands, such as
VIEWSTATE and VIEWSAVE.
In addition to state modification via communication with the IRIT
server, modes can be interactively modified on most of the display devices
using a pop-up menu that is activated using the right button in the
transformation window}.
This pop up menu is somewhat different in different drivers, but its
entries closely follow the entries of the above state command table.
All the display drivers are now able to animate objects with animation
curves' attributes on them. For more on the way animation curves can be
created see the Animation Section of this manual
Once a scene with animation curves' attributes is being loaded into
a display device, one can enter "animation" mode using the "Animation"
button available in all display devices. The user is then prompt (either
graphically or in a textual based interface) for the starting time,
termination time and step size of the animation. The parameter space of
the animation curve is serving as the time domain. The default staring
and terminating times are set as the minimal and maximal parametric
domain values of all animation curves. An object at time t below the
minimal parametric value will be placed at the starting value of the
animation curve. Similarly, an object at time t above the
maximal parametric value will be placed at the termination value of the
animation curve. The user can also set a bouncing back and forth mode,
the number of repetitions, and if desired, request the saving of all
the different scenes in the animation as seperate files so a high quality
animation can be created.
The x11drvs supports the following X Defaults (searched at ~/.Xdefaults):
#ifndef COLOR
irit*MaxColors: 1
irit*Trans*BackGround: Black
irit*Trans*BorderColor: White
irit*Trans*TextColor: White
irit*Trans*SubWin*BackGround: Black
irit*Trans*SubWin*BorderColor: White
irit*Trans*CursorColor: White
irit*View*BackGround: Black
irit*View*BorderColor: White
irit*View*CursorColor: White
#else
irit*MaxColors: 15
irit*Trans*BackGround: NavyBlue
irit*Trans*BorderColor: Red
irit*Trans*TextColor: Yellow
irit*Trans*SubWin*BackGround: DarkGreen
irit*Trans*SubWin*BorderColor: Magenta
irit*Trans*CursorColor: Green
irit*View*BackGround: NavyBlue
irit*View*BorderColor: Red
irit*View*CursorColor: Red
#endif
irit*Trans*BorderWidth: 3
irit*Trans*Geometry: =150x500+510+0
irit*View*BorderWidth: 3
irit*View*Geometry: =500x500+0+0
The Motif-based display drivers contain three types of gadgets which can be
operated in the following manner. Scales: can be dragged or clicked outside
button for single (mouse's middle button) or continuous (mouse's left
button) action. Pushbuttons: activated by clicking the mouse's left button.
The control panel: allowes rotation, translation of the objects in three
axes, determine perspective ratio, viewing object from top, side, front or
isometrically, determining scale factor and clipping settings, and operate
the matrix stack.
The environment window toggles between screen or object transformation,
depth cue on or off, orthographic or perspective projection, wireframe or
solid display, single or double buffering, showing or hiding normals,
including or excluding the surface;s mesh and curve;s control polygon,
surface drawing using isolines or polygons, and four or two polygons per
flat patch. Same display devices allow for the inclusion or exclusion of
internal edges, and enable or disable of antialiased lines.
Scales in the X11/Motif based devices set normals length, lines width,
control sensitivity, the number of islolines and samples, etc.
The locations of windows as set via [-g] and [-G] and/or via the
configuration file overwrites in x11drvs the Geometry X11 defaults.
To use the Geometry X11 default use '-G " "' and '-g " "' or set the
string to empty size in the configuration file.
In os2drvs, only -G is used to specify the dimensions of the parent window
that holds both the viewing and the transformation window.
In os2drvs, the following key strokes are available as short cuts:
Key Function
^x Quit
^s Save
^f Front View
^d Side View
^t Top View
^i Isometric VIew
^p Perspetive/Orthographic
^n View Internal Edges
^v View Vertices' Normals
^g View Polygons' Normals
^b Backface Culling
^c Depth Cue
^m View Control Mesh/Poly
xglmdrvs -z
prints all the options and their current values.
xglmdrvs -B -i -l 3 solid1.dat
displays the model of solid1.dat using back face culling ('-B'), with
internal edges ('-i'), and line width of 3.
xglmdrvs -r -A flat wiggle.dat
displays the model of wiggle.dat shaded ('-r') using flat shading ('-A').
xglmdrvs -I 40 -u -b 255 255 255 wiggle.dat
displays the model of wiggle.dat using isolines' density of 40 ('-I'), using
unit matrix to begin with ('-u'), and a white background ('-b').
xglmdrvs -X 0,2,0.1,sx -r anim.dat
executes the animation in anim.dat, from time 0 to time 2 in steps of 0.1.
The animation is saved in one frame per file (flag 's' in '-X') and
the display device exists once the animation has terminated (flag 'x'
in '-X')). The animation will be shaded ('-r').
$
Utilities - General Usage
The IRIT solid modeler is accompanied by quite a few utilities. They
can be subdivided into two major groups. The first includes auxiliary tools
such as illustrt and poly3d-h. The second includes filters such as irit2ray
and irit2ps.
All these tools operate on input files, and most of the time produce
data files. In all utilities that read files, the dash ('-') can be used
to read stdin.
Example:
poly3d-h solid1.dat | irit2ps - > solid1.ps
All the utilities have command line options. If an option is set by a '-x'
then '-x-' resets the option. The command line options overwrite the settings
in config files, and the reset option is useful for cases where the option
is set by default, in the configuration file.
All utilities can read a sequence of data files. However, the last
transformation matrices found (VIEW_MAT and PRSP_MAT) are actually used.
Example:
poly3d-h solid1.dat | x11drvs solid1.dat - solid1.mat
x11drvs will display the original solid1.dat file with its hidden version,
as computed by poly3d-h, all with the solid1.mat, ignoring all other matrices
in the data stream.
Under unix, compressed files with a postfix ".Z" will be automatically
uncompressed on read and write. The following is legal under unix,
poly3d-h solid1.dat.Z | x11drvs solid1.dat.Z - solid1.mat
where solid1.dat.Z was saved from within IRIT using the command
save( "solid1.dat.Z", solid1 );
or similar. The unix system's "compress" and "zcat" are used for the purpose
of (un)compressing the data via pipes. See also SAVE and LOAD.
$
Poly3d-h - Hidden Line Removing Program
poly3d-h is a program to remove hidden lines from a given polygonal model.
Freeform objects are preprocessed into polygons with controlled fineness.
The program performs 4 passes over the input:
1. Preprocesses and maps all polygons in a scene, and sorts them.
2. Generates edges out of the polygonal model and sorts them (preprocessing
for the scan line algorithm) into buckets.
3. Intersects edges, and splits edges with non-homogeneous visibility (the
scan line algorithm).
4. Applies a visibility test of each edge.
This program can handle CONVEX polygons only. From IRIT one can
ensure that a model consists of convex polygons only using the CONVEX command:
CnvxObj = convex( Obj );
just before saving it into a file. Surfaces are always decomposed into
triangles.
poly3d-h output is in the form of polylines. It is a regular IRIT data
file that can be viewed using any of the display devices, for example.
poly3d-h [-b] [-m] [-i] [-e #Edges] [-H] [-4] [-W Width]
[-F PolyOpti FineNess] [-q] [-o OutName] [-t AnimTime]
[-c] [-z] DFiles > OutFile
-b: BackFacing - if an object is closed (such as most models created by
IRIT), back facing polygons can be deleted, therefore speeding up
the process by at least a factor of two.
-m: More - provides some more information on the data file(s) parsed.
-i: Internal edges (created by IRIT) - default is not to
display them, and this option will force displaying them as well.
-e n: Number of edges to use from each given polygon (default all).
Handy as '-e 1 -4' for freeform data.
-H: Dumps both visible lines and hidden lines as separated objects.
Hidden lines will be dumped using a different (dimmer) color and (a
narrower) line width.
-4: Forces four polygons per almost flat region in the surface to
polygon conversion. Otherwise two polygons only.
-W Width: Selects a default width for visible lines in inches.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-q: Quiet mode. No printing aside from fatal errors. Disables -m.
-o OutName: Name of output file. Default is stdout.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-z: Prints version number and current defaults.
-c: Clips data to screen (default). If disabled ('-c-'), data
outside the view screen ([-1, 1] in x and y) are also processed.
Some of the options may be turned on in poly3d-h.cfg. They can be then
turned off in the command line as '-?-'.
The program can be configured using a configuration file named poly3d-h.cfg.
This is a plain ASCII file you can edit directly and set the parameters
according to the comments there. 'poly3d-h -z' will display the current
configuration as read from the configuration file.
The configuration file is searched in the directory specified by the
IRIT_PATH environment variable. For example,
'setenv IRIT_PATH /u/gershon/irit/bin/'.
If the IRIT_PATH variable is not set, the current directory is searched.
As this program is not interactive, usage is quite simple, and the only
control available is using the command line options.
$
If a certain surface should be polygonized into a finer/caurser
set of polygons than the rest of the scene, one can set a "resolution"
attribute which specifies the relative FineNess resolution of this
specific surface. Further, "u_resolution" and "v_resolution" might be
similarly used to set relative resolution for the u or v direction only.
Poly3d-r - A Simple Data Rendering Program
Retired. Sources can be found in the contrib directory, but this program is
no longer supported. See irender program instead.
$
Illustrt - Simple line illustration filter
illustrt is a filter that processes IRIT data files and dumps out modified
IRIT data files. illustrt can be used to make simple nice illustrations of
data. The features of illustrt include depth sorting, hidden line clipping
at intersection points, and vertex enhancements. illustrt is designed to
closely interact with irit2ps, although it is not neceessary to use irit2ps
on illustrt output.
illustrt [-I #IsoLines] [-f PolyOpti SampTol] [-s] [-M] [-P] [-p] [-O]
[-l MaxLnLen] [-a] [-t TrimInter] [-o OutName] [-Z InterSameZ]
[-m] [-T AnimTime] [-z] DFiles
-I #IsoLines: Specifies number of isolines per surface, per direction.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-s: sorts the data in Z depth order that emulates hidden line removal
once the data are drawn.
-M: Dumps the control mesh/polygon as well.
-P: Dumps the curve/surface as isocurves.
-p: Dumps vertices of polygons/lines as points.
-O: Handle polygonal objects as possibly open. This will generate
two identical edges for an edge shared by two adjacent polygons.
Can be useful for open or isilated polygons.
-l MaxLnLen: breaks long lines into shorter ones with maximal length
of MaxLnLen. This option is necessary to achieve good depth depending
line width in the '-d' option of irit2ps.
-a: takes into account the angle between the two (poly)lines that
intersect when computing how much to trim. See also -t.
-t TrimInter: Each time two (poly)line segments intersect in the
projection plane, the (poly)line that is farther away from the
viewer is clipped TrimInter amount from both sides. See also -a.
-o OutName: Name of output file. Default is stdout.
-Z InterSameZ: The maximal Z depth difference of intersection curves
to be be considered invalid.
-m: More talkative mode. Prints processing information.
-T AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-z: Prints version number and current defaults.
illustrt is a simple line illustration tool. It process geometry such
as polylines and surfaces and dumps geometry with attributes that will make
nice line illustrations. illustrt is geared mainly toward its use with
irit2ps to create postscript illustrations. Here is a simple example:
illustrt -s -l 0.1 solid1.dat | irit2ps -W 0.05 -d 0.2 0.6 -u - > solid.ps
make sure all segments piped into irit2ps are shorter than 0.1 and sort them
in order to make sure hidden surface removal is correctly applied. Irit2ps
is invoked with depth cueing activated, and a default width of 0.05.
illustrt dumps out regular IRIT data files, so output can be handled
like any other data set. illustrt does the following processing to the input
data set:
Converts surfaces to isocurves ('-I' flag) and isocurves and curves to
polylines ('-S' flag), and converts polygons to polylines.
Polygonal objects are considered closed and even though each edge is
shared by two polygons, only a single one is generated.
Finds the intersection location in the projection plane of all segments in
the input data set and trims away the far segment at both sides of the
intersection point by an amount controlled by the '-t' and '-a' flags.
Breaks polylines and long lines into short segments, as specified via the
'-l' flag, so that width depth cueing can be applied more accurately
(see irit2ps's '-d' flag) as well as the Z sorting.
Generates vertices of polygons in the input data set as points in output data
controlled via the '-p' flag.
set.
Applies a Z sort to the output data, if '-s', so drawing in order of the data
will produce a properly hidden surface removal drawing.
Here is a more complex example. Make sure tubular is properly set via
"attrib(solid1, "tubular", 0.7);" and invoke:
illustrt -s -p -l 0.1 -t 0.05 solid1.dat |
irit2ps -W 0.05 -d 0.2 0.6 -p h 0.05 -u - > solid.ps
makes sure all segments piped into irit2ps are shorter than 0.1, generates
points for the vertices, sorts the data in order to make sure hidden surface
removal is correctly applied, and trims the far edge by 0.05 at an
intersection point. Irit2ps is invoked with depth cueing activated and a
default width of 0.05, points are drawn as hollowed circles of default
size 0.05, and lines are drawn tubular.
Objects in the input stream that have an attribute by the name of
"IllustrtNoProcess" are passed to the output unmodified.
Objects in the input stream that have an attribute by the name of
"SpeedWave" will have a linear segments added that emulate fast
motion with the following attributes,
"Randomness,DirX,DirY,DirZ,Len,Dist,LenRandom,DistRandom,Width".
Objects in the input stream that have an attribute by the name of
"HeatWave" will have a spiral curves added that emulate a heat wave
in the +Z axis with the following attributes,
"Randomness,Len,Dist,LenRandom,DistRandom,Width".
Examples:
attrib(Axis, "IllustrtNoProcess", "");
attrib(Obj, "SpeedWave", "0.0005,1,0,0,5,3,3,2,0.05");
attrib(Obj, "HeatWave", "0.015,0.1,0.03,0.06,0.03,0.002");
$
Irender - Simple Scan Line Renderer
irender is a program to render IRIT scenes into images. It is
a software based Z buffer that is able to create images in few formats.
Several of its features includes parametric and volumetric texture mapping,
shadow computations, transparency and antialiasing.
Freeform objects are preprocessed into polygons with controlled fineness.
irender [-z] [-v] [-s XSize YSize] [-a Ambient] [-b R G B] [-B]
[-F PolyOpti FineNess] [-f PolyOpti SampTol]
[-M Flat/Gouraud/Phong] [-P WMin [WMax]] [-S] [-T] [-t AnimTime]
[-A FilterName] [-Z] [-n] [-i rle/ppm] files
-z: Prints version number and current defaults.
-v: Verbose mode. Prints informative messages as it progresses.
-s XSize YSize: Sets the size of the output image, in pixels.
Default to 512x512.
-a Ambient: Sets the ambient lighting fraction. Between zero (no
ambient lighting) and one. Default to 0.2.
-b R G B: Sets the background color. Each of thre R,G,B colors is
an integer value between zero and 255. Default to black.
-B : Apply back face culling. Somewhat faster, but only correct for
closed objects. Default is no back face culling.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. Default is 0 and 20.0 (no optimal sampling with fineness
of 20.0 (real number)).
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-M Flat/Gouraud/Phong: Selects the shader to be used. Default to
Phong if has normals of vertices, Flat if no normals are found.
-P WMin [WMax]: Width of rendered polyline, in pixels. If only
WMin is specified, all polylines are set to have WMin width. Otherwise,
if WMax is prescribed as well, polylines' width is set to be
proportional to their depth with WMax is the width of closest polyline
and WMin the farest polyline.
-S: Enable shadow computation. No shadows will be rendered without -S.
-T: Enable transparency computation. No transparent object will be
processed without -T.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-A FilterName: Selects an antialiasing filter. FilterName can be one
of 'none', 'box', 'triangle', 'quadratic', 'cubic', 'catrom',
'mitchell', 'gaussian', 'sinc, and 'bessel'. Default is 'none'.
-Z: Output will be in the form of Z depth instead of a color image.
Output will be 32 bits depth instead of RGBA.
-n: Reverses the normals of vertices and planes, globally.
-i rle/ppm: Selects output image type. Currently the Utah Raster
Toolkit's (URT) rle format is being supported as well as the PPM
format.
Some of the options may be turned on in irender.cfg. They can be then
turned off in the command line as '-?-'.
The program can be configured using a configuration file named irender.cfg.
This is a plain ASCII file you can edit directly and set the parameters
according to the comments there. 'irender -z' will display the current
configuration as read from the configuration file.
The configuration file is searched in the directory specified by the
IRIT_PATH environment variable. For example,
'setenv IRIT_PATH /u/gershon/irit/bin/'.
If the IRIT_PATH variable is not set, the current directory is searched.
As this program is not interactive, usage is quite simple, and the only
control available is using the command line options.
One can specify several attributes that affect the way the scene is
rendered. The attributes can be generated within IRIT.
See also the ATTRIB IRIT command.
Surface color is controlled in two levels. If the object has an RGB
attribute, it is used. Otherwise, a color as set via the IRIT COLOR
command is used.
If a certain surface should be finer/caurser than the rest of the
scene, one can set a "resolution" attribute which specifies the
relative FineNess resolution of this specific surface. Further,
"u_resolution" and "v_resolution" might be similarly used to set
relative resolution for the u or v direction only.
Example:
attrib( Ball, "rgb", "255,0,0" );
color( Sphere, white );
The cosine exponent of the phong shader can be set for a specific
object via the SRF_COSINE attribute.
Example:
attrib( Ball, "srf_cosine", 16 );
An object can be drawn transparent instead of opaque, if it has a
"transp" attribute. A transparent value of one denotes a completely
transparent object, while a value of zero means a completely opaque
object. Transparent object will be rendered as such if and only if the '-T'
command line option is set.
Example:
attrib( final, "transp", 0.5 );
Several types of texture mapping are supported. Parametric texture may be
attached to a parametric surface where the prescribed image is mapped onto
the rectangular parametric domain of the surface.
Example:
attrib( Srf1, "ptexture", "checker.ppm" );
The program will automatically detected a ppm file from an rle according to
the file's name.
A second type of texture mapping can be applied to all geometric objects.
Herein, a procedural texture mapping is employed. Current supported
textures are:
wood Wood style
marble Marble style
contour Parallel plane contouring
ncontour Constant normal angle to major axis
A second parameter that must be provided for procedural textures is the
scaling factor of the texture, which can be either one parameter of uniform
scaling or a vector of three coefficients for scaling in x, y, and z. The
two parameters are separated by a comma. For contour style, the scale
denotes the spacing of the contouring planes in X, Y and Z. For ncontour
style, the scale also denotes the spacing of the adjacent constant normal
contours. Related attributes are "texture_color" and "texture_width" that
supports the color and the width of the textured strokes.
Example:
attrib( Obj1, "texture", "marble, 2" );
attrib( Obj2, "texture", "contour, 1 0.5 2.5" );
attrib( Obj2, "texture_width", "0.05" );
attrib( Obj2, "texture_color", "255,255,255" );
which sets Obj1 to have a marble procedural texture with a uniform
scaling factor of 2 and a contouring texture for Obj2 with scaling
factors of (1, 0.5, 2.5) in x, y, and z, in white color and width 0.05.
In addition, a scalar surface spanning the same parameteric domain as an
original surface may be used as texture mapping function. Herein, the
scalar function texture is evaluated at each UV parameter value and is mapped
through a color scale to yield the output color. This type of texture is
useful for stress maps or analysis maps on top of freeform surfaces.
Several related attributes are supported: "stexture_scale" which prescribes
the color scale image (only its first column is employed), and
"stexture_bound" that sets the domain that will be clipped to the min max
values. Funally, "stexture_func" can hold the functions "sqrt" or "abs"
to be applied to the evaluated surface value.
Example:
attrib( Srf, "stexture", scrvtr( Srf, P1, off ) );
attrib( Srf, "stexture_scale", "color_scale.ppm" );
attrib( Srf, "stexture_func", "sqrt" );
attrib( Srf, "stexture_bound", "0.0 100.0" );
where scrvtr computes a scalar field to Srf that represents the
sum of the squares of the principle curvatures. The evaluated scalar
texture surface's value is piped through a sqrt function.
The first column of the image of color_scale.ppm is used to set
the coloring scale for curvature bounds values between 0.0 and 100.0.
Both "stexture_scale" and "stexture_bound" are optional. The default
color scale maps the min/max values from blue to red through green. The
default scalar surface texture bound is computed as the extreme values of
the "stexture" surface.
While the program has a default for lighting which is two light sources
at opposite directions at (1, 1, 1) and (-1, -1, -1), one can overwrite
this default. A POINT_TYPE object with LIGHT_SOURCE attribute denotes
a light source. If irender detects one or more light sources in the
input stream, the default light sources are not created. Two types of
light sources may be prescribed, a parallel at infinity or a point at
finite distance light source, distinguished by a TYPE attribute of
either POINT_POS or POINT_INFTY. A point light source can be colored,
when an RGB attribute will set its color. A point light source will cast
shadows if and only if it has SHADOW attribute (one needs to apply the '-S'
command line option as well for rendering shadows). Finally, one can
construct two mirrored light sources at opposite directions if TWOLIGHT
attribute is added to the light source object.
Example:
Light1 = point( 0, 0, 10 );
attrib( Light1, "light_source", on );
attrib( Light1, "shadow", on );
attrib( Light1, "rgb", "255,0,0" );
attrib( Light1, "type", "point_pos" );
Light2 = point( 1, 1, 1 );
attrib( Light2, "light_source", on );
attrib( Light2, "twolight", on );
attrib( Light2, "type", "point_infty" );
constructs two lights sources with Light1 with red color positioned
at (0, 0, 10) and casting shadows, while Light2 will create two
mirrored white parallel lights sources in the direction of (1, 1, 1) and
(-1, -1, -1), as is irender's default.
$
DAT2BIN - Data To Binary Data file filter
dat2bin [-t] [-z] DFiles
-t: Dumps data to stdout as text instead of binary.
-z: Print version number and current defaults.
It may be sometimes desired to convert .dat data files into a binary form,
for example, for fast loading of files with large geometry. Binary files can
be somewhat larger, are unreadable in editors but are much faster to load in.
A binary file must have a '.bdt' file type.
Example:
dat2bin b58polys.dat > b58polys.bdt
dat2bin -t b58polys.bdt | more
to convert a text file b58polys.dat into a binary file b58polys.bdt and
to view the content of the binary file by converting it back to text. At
this time data through pipes must be in text. That is, the following is
illegal:
dat2bin b58polys.dat | xglmdrvs -
It should be remembered that the binary format is not documented and
it might change in the future. Moreover, it is machine dependent and can
very well may be unreadible between different platforms.
$
DAT2IRIT - Data To IRIT file filter
Converts '.dat' and '.bdt' data files to '.irt' IRIT scripts.
dat2irit [-z] DFiles
-z: Print version number and current defaults.
It may be sometimes desired to convert .dat data files into a form that
can be fed back to IRIT - a '.irt' file. This filter does exactly that.
Example:
dat2irit b58.dat > b58-new.irt
$
DXF2IRIT - DXF (Autocad) To IRIT filter
Converts Autocad's, DXF data files into IRIT data files.
dxf2irit [-m] [-f] [-o OutName] [-z] DXFFile
-m: Provides some more information on the data file(s) parsed.
-f: Coerce floating end conditions to constructed freeform surfaces.
Default is open end conditions.
-o OutName: Name of output file. By default the output goes to
stdout.
-z: Prints version number and current defaults.
dxf2irit converts Autocad's DXF data files into IRIT data files. The
current version provides only partial support for the conversion of
freeform surfaces, mainly due to luck of documentation examples on the
dxf format and the convoluted way freeform surfaces are saved.
Example:
dxf2irit file.dxf > file.dat
$
IRIT2DXF - IRIT To DXF (Autocad) filter
Converts IRIT data files into Autocad's, DXF data files.
irit2dxf [-s Scale] [-t Tx Ty Tz] [-i] [-f] [-F PolyOpti FineNess]
[-4] [-o OutName] [-T] [-a AnimTime] [-z] DFiles
-s Scale: Global scaling factor of the converted geometry.
-t Tx Ty Tz: a Vector of size three of translation factors along
the X, Y, and Z axes.
-i: Show internal edges as well.
-f: Dumps freeforms as converted polygonal geometry.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-4: Forces four polygons per almost flat region in the surface to
polygon conversion. Otherwise two polygons only.
-o OutName: Name of output file. By default the output goes to
stdout.
-T: Talkative mode. Prints processing information.
-a AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-z: Prints version number and current defaults.
irit2dxf converts IRIT data files into Autocad's DXF data files. The
current version provides only partial support for the direct conversion of
freeform surfaces, mainly due to luck of documentation examples on the
dxf format and the convoluted way freeform surfaces are saved. Nonetheless,
Freeform surfaces can be converted into polygons using the '-f' flag.
Example:
irit2dxf -z -t 1 2 3 -F 0 20 -4 -o file.dxf file.dat
$
IRIT2Hgl - IRIT To HPGL filter
Converts IRIT geometry into the HL Graphics Language used by HP's plotters.
irit2hgl [-t XTrans YTrans] [-I #UIso[:#VIso[:#WIso]]]
[-f PolyOpti SampTol] [-F PolyOpti FineNess] [-M] [-G] [-T]
[-a AnimTime] [-i] [-o OutName] [-z] DFiles
-t XTrans YTrans: X and Y translation. of the image. Default is (0, 0).
-I #UIso[:#VIso]: Specifies the number of isolines per surface, per
direction. If #VIso is not specified, #UIso is used for #VIso as
well.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-F PolygonOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4. This enforces the dump of freefrom geometry
as polygons.
-M: Dumps the control mesh/polygon as well.
-G: Dumps the freeform geometry.
-T: Talkative mode. Prints processing information.
-a AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-i: Internal edges (created by IRIT) - default is not to
display them, and this option will force displaying them as well.
-o OutName: Name of output file. By default the name of the first data
file from DFiles list is used. See below on the output files.
-z: Prints version number and current defaults.
Irit2Hgl converts freeform surfaces and polygons into polylines in a format
that can be used by HPGL.
Example:
irit2Hgl -M -f 0 16 saddle.dat > saddle.hgl
However, one can overwrite the viewing matrix by appending a new matrix
in the end of the command line, created by the display devices:
x11drvs b58.dat
irit2Hgl -M -f 0 16 b58.dat irit.mat > saddle.hgl
where irit.mat is the viewing matrix created by x11drvs.
$
IRIT2IV - IRIT To SGI's Inventor filter
IV is the format used by the Inventor modeling/rendering package from SGI.
irit2iv [-l] [-4] [-P] [-F PolyOpti FineNess] [-f PolyOpti SampTol]
[-T] [t AnimTime] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although, most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free form
shape (saddle like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-P: Polygonize freeform shapes. Default is to leave freeform curves
and surfaces as is.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-z: Prints version number and current defaults.
Irit2Iv converts freeform surfaces and polygons into polygons and
saved in iv Inventor's ascii file format.
Example:
irit2iv solid1.dat > solid1.iv
Surfaces are converted to polygons with fineness control:
irit2iv -F 0 16 - view.mat < saddle.dat > saddle.iv
Note the use of '-' for stdin.
$
IRIT2NFF - IRIT To NFF filter
irit2nff [-l] [-4] [-c] [-F PolyOpti FineNess] [-o OutName] [-T]
[-t AnimTime] [-g] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although, most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free-form
shape (saddle-like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-c: Output files should be filtered by cpp. When set, the usually
huge geometry file is separated from the main nff file that contains
the surface properties and view parameters. By default all data,
including the geometry, are saved into a single file with type extension
'.nff'. Use of '-c' will pull out all the geometry into a file with
the same name but a '.geom' extension, which will be included using the
'#include' command. The '.nff' file should, in that case, be
preprocessed using cpp before being piped into the nff renderer.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-o OutName: Name of output file. By default the name of the first data
file from the DFiles list is used. See below on the output files.
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-g: Generates the geometry file only. See below.
-z: Prints version number and current defaults.
Irit2Nff converts freeform surfaces into polygons in a format that can
be used by an NFF renderer. Usually, one file is created with '.nff' type
extension. Since the number of polygons can be extremely large, a '-c'
option is provided, which separates the geometry from the surface
properties and view specification, but requires preprocessing by cpp.
The geometry is isolated in a file with extension '.geom' and included
(via '#include') in the main '.nff' file. The latter holds the surface
properties for all the geometry as well as the viewing specification.
This allows for the changing of shading or the viewing properties while
editing small ('.nff') files.
If '-g' is specified, only the '.geom' file is created, preserving the
current '.nff' file. The '-g' flag can be specified only with '-c'.
In practice, it may be useful to create a low resolution approximation
of the model, change viewing/shading parameters in the '.nff' file until
a good view and/or surface quality is found, and then run Irit2Nff once more
to create a high resolution approximation of the geometry using '-g'.
Example:
irit2nff -c -l -F 0 8 b58.dat
creates b58.nff and b58.geom with low resolution (FineNess of 5).
Once done with parameter setting, a fine approximation of the model can
be created with:
irit2nff -c -l -g -F 0 64 b58.dat
which will only recreate b58.geom (because of the -g option).
One can overwrite the viewing matrix by appending a new matrix in the
end of the command line, created by a display device:
xgldrvs b58.dat
irit2nff -l -F 0 32 b58.dat irit.mat
where irit.mat is the viewing matrix created by xgldrvs.
One can specify surface qualities for individual surfaces of a model.
Several such attributes are supported by Irit2Nff and can be set within
IRIT. See also the ATTRIB IRIT command.
If a certain surface should be finer/caurser than the rest of the
scene, one can set a "resolution" attribute which specifies the
relative FineNess resolution of this specific surface. Further,
"u_resolution" and "v_resolution" might be similarly used to set
relative resolution for the u or v direction only.
Example:
attrib( srf1, "resolution", 2 );
will force srf1 to have twice the default resolution, as set via the '-f'
flag.
Almost flat patches are converted to polygons. The rectangle can be
converted into two polygons (by subdividing along one of its diagonals) or
into four by introducing a new point at the center of the patch. This
behavior is controlled by the '-4' flag, but can be overwritten for
individual surfaces by setting a "twoperflat" or a "fourperflat" attribute.
NFF specific properties are controlled via the following attributes:
"kd", "ks", "shine", "trans", "index". Refer to the NFF manual for detail.
Example:
attrib( srf1, "kd", 0.3 );
attrib( srf1, "shine", 30 );
Surface color is controlled in two levels. If the object has an RGB
attribute, it is used. Otherwise, a color, as set via the IRIT COLOR
command, is used if set.
Example:
attrib( tankBody, "rgb", "244,164,96" );
$
IRIT2PLG - IRIT To PLG (REND386) filter
PLG is the format used by the rend386 real time renderer for the IBM PC.
irit2plg [-l] [-4] [-F PolyOpti FineNess] [-T] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although, most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free form
shape (saddle like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-T: Talkative mode. Prints processing information.
-z: Prints version number and current defaults.
Irit2Plg converts freeform surfaces and polygons into polygons in a
format that can be used by the REND386 renderer.
Example:
irit2plg solid1.dat > solid1.plg
Surfaces are converted to polygons with fineness control:
irit2plg -F 0 16 - view.mat < saddle.dat > saddle.plg
Note the use of '-' for stdin.
$
IRIT2POV - IRIT To POVRAY raytracer filter
irit2pov [-l] [-4] [-C] [-F PolyOpti FineNess] [-f PolyOpti SampTol]
[-o OutName] [-g] [-p Zmin Zmax] [-P] [-M] [-T] [-t AnimTime]
[-I #UIso[:#VIso[:#WIso]]] [-s ObjSeq#] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free-form
shape (saddle-like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-C: Construct biCubic Bezier patches whenever possible as POVRAY
support this type of surface. Polynomial Bezier surfaces of
orders up to and including bicubic (order 4, degree 3) are degree
raised to bicubic. Piecewise polynomials Bspline surfaces are
split into Bezier patches. Higher order surfaces and rational
surfaces are always converted into polygons.
-F PolygonOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4, -C, and -l.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-o OutName: Name of output file. By default the name of the first data
file from the DFiles list is used. See below on the output
files.
-g: Generates the geometry file only. See below.
-p Zmin Zmax: Sets the ratios between the depth cue and the width of
the dumped polylines. See also -P. Closer lines will be drawn
wider.
-P: Forces dumping polygons as polylines with thickness controlled
by -p.
-M: If -P (see -P and -p) then converts the control mesh/polygon
to polylines which are represented as a sequence of truncated
cones.
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-I #UIso[:#VIso[:#WIso]]: Specifies the number of isolines per
surface/trivariate, per direction. If #VIso or #WIso is not
specified, #UIso is used for #VIso etc.
-s ObjSeq#: Sets object sequence number if no object name. Default 1.
-z: Prints version number and current defaults.
Irit2pov converts freeform surfaces into polygons in a format that can
be used by the POVRAY ray tracing program.
Two files are created, one with a '.geom' extension and
one with '.pov'. Since the number of polygons can be extremely large,
the geometry is isolated in the '.geom' file and is included
(via '#include') in the main '.pov' file. The latter holds the surface
properties for all the geometry as well as viewing and POVRAY specific
commands. This allows for the changing of the shading or the viewing
properties while editing small ('.pov') files.
If '-g' is specified, only the '.geom' file is created, preserving the
current, possibly manually modified, '.pov' file.
In practice, it may be useful to create a low resolution approximation
of the model, change the viewing/shading parameters in the '.pov' file until
a good view and/or surface quality is found, and then run Irit2pov once more
to create a high resolution approximation of the geometry using '-g'.
Example:
irit2pov -l -F 0 5 b58.dat
creates b58.pov and b58.geom with low resolution (FineNess of 5).
At such low resolution it can very well may happen that triangles will have
normals "over the edge" since a single polygon may approximate a highly
curved surface.
One can ray trace this scene using a command similar to:
POVRAY -Q0 +Ib58
Once done with parameter setting for POVRAY, a fine approximation of the
model can be created with:
irit2pov -l -g -F 0 64 b58.dat
which will only recreate b58.geom (because of the -g option).
Interesting effects can be created using the depth cue support and polyline
conversion of irit2pov. For example
irit2pov -P -p -0.0 0.5 solid1.dat
will dump solid1 as a set of polylines (represented as truncated cones in
POVRAY) with varying thickness according to the z depth. Another example
is
irit2pov -P -p -0.1 1.0 saddle.dat
which dumps the isolines extracted from the saddle surface with varying
thickness.
Each time a data file is saved in IRIT, it can be saved with the
viewing matrix of the last INTERACT by saving the VIEW_MAT object as well.
I.e.:
save( "b58", b58 );
However one can overwrite the viewing matrix by appending a new matrix
in the end of the command line, created by the display devices:
xglmdrvs b58.dat // Creates irit.mat
irit2pov -l -F 0 16 b58.dat irit.mat
where irit.mat is the viewing matrix created by xglmdrvs. The output name,
by default, is the last input file name, so you might want to provide an
explicit name with the -o flag.
One can specify surface qualities for individual surfaces of a model.
Several such attributes are supported by Irit2pov and can be set within
IRIT. See also the ATTRIB IRIT command.
If a certain surface should be finer/caurser than the rest of the
scene, one can set a "resolution" attribute which specifies the
relative FineNess resolution of this specific surface. Further,
"u_resolution" and "v_resolution" might be similarly used to set
relative resolution for the u or v direction only.
Example:
attrib( srf1, "resolution", 2 );
will force srf1 to have twice the default resolution, as set via the '-f'
flag.
Almost flat patches are converted to polygons. The rectangle can be
converted into two polygons (by subdividing along one of its diagonals) or
into four by introducing a new point at the patch center. This behavior is
controlled by the '-4' flag, but can be overwritten for individual surfaces
bu setting "twoperflat" or "fourperflat".
POVRAY also supports bicubic Bezier patches and the '-C' option of
irit2pov supports that. In such a case, the resolution that is requested
from POVRAY to polygonize these patches approximately follow the resolution
as selected via the '-F' flag of irit2pov. Nevertheless, one can
override the requested resolution via the "steps", "u_steps", and "v_steps"
attributes to irit2pov data files that are transfered directly to
POVRAY's bicubic Bezier patches. The "steps" attributes sets both
"u_steps" and "v_steps".
POVRAY specific properties are controlled via the following attributes:
"ambient", "diffuse", "brilliance", "phong", "phong_size", "specular",
"roughness", "metallic", "reflection", "refraction", "ior", "caustics",
"fade_distance", "fade_power", "irid", "crand", "texture", "pigment",
"finish", "halo", and "normal. The value of this attributes must be
strings as it is copied verbatim. Refer to POVRAY's manual for their
exact meaning.
Example:
attrib( legs, "ambient", "0.1" );
attrib( pot, "matallic", "" );
attrib( table, "ior", "1.4" );
Surface color is controlled in two levels. If the object has an RGB
attribute, it is used. Otherwise a color as set via the IRIT COLOR
command is being used if set.
Example:
attrib( tankBody, "rgb", "244,164,96" );
$
IRIT2PS - IRIT To PS filter
irit2ps [-l] [-4] [-s Size] [-I #UIso[:#VIso[:#WIso]]] [-F PolyOpti FineNess]
[-f PolyOpti SampTol] [-M] [-G] [-P] [-W LineWidth]
[-w WidenLen WidenWidth] [-b R G B] [-B X1 Y1 X2 Y2] [-c] [-C]
[-T] [-t AnimTime] [-i] [-o OutName] [-d [Zmin Zmax]]
[-D [Zmin Zmax]] [-p PtType PtSize] [-u] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free-form
shape (saddle-like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-s Size: Controls the size of the postscript output in inches.
Default is to fill the entire screen.
-I #UIso[:#VIso[:#WIso]]: Specifies the number of isolines per
surface/trivariate, per direction. If #VIso or #WIso is not
specified, #UIso is used for #VIso etc.
-F PolygonOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-M: Dumps the control mesh/polygon as well.
-G: Dumps the curve/surface (as freeform geometry). Default. See -I,
-C, -f for control on polyline approximation.
-P: Dumps the curve/surface (as polygons). See -F, -l, -4 for control
on polygonal approximation.
-W #LineWidth: Sets the line drawing width in inches. Default is
as thin as possible. This option will overwrite only those objects
that do not have a "width" attribute. See also -d. If LineWidth
is negative its absolute value is used to scale the current width
of the object if has one, or the default width otherwise.
-w WidenLen WidenWidth: If end points of polylines should be made
wider, and if so to what width.
-b R G B: Sets a colored background. RGB are three integers prescribing
the Red, Green, and Blue coefficients. if no -c (i.e. a gray level
drawing) this color is converted to a gray level using RGB to
T.V. Y(IQ) channel conversion.
-B X1 Y1 X2 Y2: Clips the drawing area outsize the bounding box
from (X1, Y1) to (X2, Y2).
-c: Creates a color postscript file.
-C: Curve mode. Dumps freeform curves and surfaces as cubic
Bezier curves. Higher order curves and surfaces and/or rationals are
approximated by cubic Bezier curves. This option generates data
files that are roughly a third of piecewise linear postscript files
(by disabling this feature, -C-), but takes a longer time to compute.
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-i: Internal edges (created by IRIT) - the default is not to
display them, and this option will force displaying them as well.
-o OutName: Name of output file. Default is stdout.
-d [Zmin Zmax]: Sets the ratios between the depth cue and the width of
the dumped data. See also -W, -p. Closer lines/points will be drawn
wider/larger. Zmin and Zmax are optional. The object's bounding
box is otherwise computed and used.
-D [Zmin Zmax]: Same as -d, but depth cue the color or gray scale
instead of width. You might need to consider the sorting option
of the illustrt tool (-s of illustrt) for proper drawings.
Only one of -d and -D can be used.
-p PtType PtSize: Specifies the way points are drawn.
PtType can be one of H, F, C for Hollow circle, Full Circle, or
Cross. PtSize specifies the size of the point to be drawn, in inches.
Vectors will also be drawn as points, but with an additional thin
line to the origin. See also -d.
-u: Forces a unit matrix transformation, i.e. no transformation.
-z: Prints version number and current defaults.
Irit2Ps converts freeform surfaces and polygons into a postscript file.
Example:
irit2ps solid1.dat > solid1.ps
Surfaces are converted to polygons with fineness control:
irit2ps -f 0 32 -c -W 0.01 saddle.dat > saddle.ps
creates a postscript file for the saddle model, in color, and with
lines 0.01 inch thick.
One can specify several attributes that affect the way the postscript
file is generated. The attributes can be generated within IRIT.
See also the ATTRIB IRIT command.
If a certain object should be thinner or thicker than the rest of the scene,
one can set a "width" attribute which specifies the line width in inches of
this specific object.
Example:
attrib( srf1, "width", 0.02 );
will force srf1 to have this width, instead of the default as set via the
'-W' flag.
If a (closed) object, a polygon for example, needs to be filled, a "fill"
attribute should be set, with a value equal to the gray level desired.
Example:
attrib( poly, "fill", 0.5 );
will fill poly with %50 gray.
If an object, a polygon for example, needs to be painted in a gray level
instead of black, a "gray" attribute should be set, with a value equal
to the gray level desired.
Example:
attrib( poly, "gray", 0.5 );
will draw poly with %50 gray.
Dotted or dashed line effects can be created using a "dash" attribute which
is a direct PostScript dash string. A simple form of this string is "[a b]"
in which a is the drawing portion (black) in inches, followed by b inches
of white space. See the postScript manual for more about the format of this
string. Here is an example for a dotted-dash line.
attrib( poly, "dash", "[0.006 0.0015 0.001 0.0015] 0" );
Surface color is controlled (for color postscript only - see -c) in two
levels. If the object has an RGB attribute, it is used. Otherwise, a color as
set via the IRIT COLOR command is used.
Example:
attrib( Ball, "rgb", "255,0,0" );
An object can be drawn as ``tubes'' instead of full lines. The ratio
between the inner and the outer radii of the tube is provided as the
TUBULAR attribute:
attrib( final, "tubular", 0.7 );
$
IRIT2RAY - IRIT To RAYSHADE filter
irit2ray [-l] [-4] [-G GridSize] [-F PolyOpti FineNess]
[-f PolyOpti SampTol] [-o OutName] [-g] [-p Zmin Zmax] [-P]
[-M] [-T] [-t AnimTime] [-I #UIso[:#VIso[:#WIso]]] [-s ObjSeq#]
[-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
by a single polygon along their linear direction.
Although most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free-form
shape (saddle-like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch. Default is 2.
-G GridSize: Usually objects are grouped as lists of polygons.
This flags will coerce the usage of the RAYSHADE grid structure,
with GridSize being used as the grid size along the object
bounding box's largest dimension.
-F PolygonOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-o OutName: Name of output file. By default the name of the first data
file from the DFiles list is used. See below on the output
files.
-g: Generates the geometry file only. See below.
-p Zmin Zmax: Sets the ratios between the depth cue and the width of
the dumped polylines. See also -P. Closer lines will be drawn
wider.
-P: Forces dumping polygons as polylines with thickness controlled
by -p.
-M: If -P (see -P and -p) then converts the control mesh/polygon
to polylines which are represented as a sequence of truncated
cones.
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-I #UIso[:#VIso[:#WIso]]: Specifies the number of isolines per
surface/trivariate, per direction. If #VIso or #WIso is not
specified, #UIso is used for #VIso etc.
-s ObjSeq#: Sets object sequence number if no object name. Default 1.
-z: Prints version number and current defaults.
Irit2Ray converts freeform surfaces into polygons in a format that can
be used by the RAYSHADE ray tracing program.
Two files are created, one with a '.geom' extension and
one with '.ray'. Since the number of polygons can be extremely large,
the geometry is isolated in the '.geom' file and is included
(via '#include') in the main '.ray' file. The latter holds the surface
properties for all the geometry as well as viewing and RAYSHADE specific
commands. This allows for the changing of the shading or the viewing
properties while editing small ('.ray') files.
If '-g' is specified, only the '.geom' file is created, preserving the
current '.ray' file.
In practice, it may be useful to create a low resolution approximation
of the model, change the viewing/shading parameters in the '.ray' file until
a good view and/or surface quality is found, and then run Irit2Ray once more
to create a high resolution approximation of the geometry using '-g'.
Example:
irit2ray -l -F 0 5 b58.dat
creates b58.ray and b58.geom with low resolution (FineNess of 5).
At such low resolution it can very well may happen that triangles will have
normals "over the edge" since a single polygon may approximate a highly
curved surface. That will cause RAYSHADE to issue an
"Inconsistent triangle normals" warning. This problem will not exist if
high fineness is used.
One can ray trace this scene using a command similar to:
RAYSHADE -p -W 256 256 b58.ray > b58.rle
Once done with parameter setting for RAYSHADE, a fine approximation of the
model can be created with:
irit2ray -l -g -F 0 64 b58.dat
which will only recreate b58.geom (because of the -g option).
Interesting effects can be created using the depth cue support and polyline
conversion of irit2ray. For example
irit2ray -G 5 -P -p -0.0 0.5 solid1.dat
will dump solid1 as a set of polylines (represented as truncated cones in
RAYSHADE) with varying thickness according to the z depth. Another example
is
irit2ray -G 5 -P -p -0.1 1.0 saddle.dat
which dumps the isolines extracted from the saddle surface with varying
thickness.
Each time a data file is saved in IRIT, it can be saved with the
viewing matrix of the last INTERACT by saving the VIEW_MAT object as well.
I.e.:
save( "b58", b58 );
However one can overwrite the viewing matrix by appending a new matrix
in the end of the command line, created by the display devices:
os2drvs b58.dat // Creates irit.mat
irit2ray -l -F 0 16 b58.dat irit.mat
where irit.mat is the viewing matrix created by os2drvs. The output name,
by default, is the last input file name, so you might want to provide an
explicit name with the -o flag.
One can specify surface qualities for individual surfaces of a model.
Several such attributes are supported by Irit2Ray and can be set within
IRIT. See also the ATTRIB IRIT command.
If a certain surface should be finer/caurser than the rest of the
scene, one can set a "resolution" attribute which specifies the
relative FineNess resolution of this specific surface. Further,
"u_resolution" and "v_resolution" might be similarly used to set
relative resolution for the u or v direction only.
Example:
attrib( srf1, "resolution", 2 );
will force srf1 to have twice the default resolution, as set via the '-f'
flag.
Almost flat patches are converted to polygons. The rectangle can be
converted into two polygons (by subdividing along one of its diagonals) or
into four by introducing a new point at the patch center. This behavior is
controlled by the '-4' flag, but can be overwritten for individual surfaces
bu setting "twoperflat" or "fourperflat".
RAYSHADE specific properties are controlled via the following attributes:
"specpow", "reflect", "transp", "body", "index", and "texture". The value of
this attributes must be strings as it is copied verbatim. Refer to
RAYSHADE's manual for their meaning.
Example:
attrib( legs, "transp", "0.3" );
attrib( legs, "texture", "wood,2" );
attrib( table, "texture", "marble" );
attrib( table, "reflect", "0.5" );
Optional scale can be prescribed to textures. In the above example
wooden legs' (that are also transparent...) texture is selected with
texture's scaling factor of 2.
Surface color is controlled in two levels. If the object has an RGB
attribute, it is used. Otherwise a color as set via the IRIT COLOR
command is being used if set.
Example:
attrib( tankBody, "rgb", "244,164,96" );
$
IRIT2Scn - IRIT To SCENE (RTrace) filter
SCENE is the format used by the RTrace ray tracer. This filter was donated
by Antonio Costa (acc@asterix.inescn.pt), the author of RTrace.
irit2scn [-l] [-4] [-F PolyOpti FineNess] [-o OutName] [-g] [-T]
[-t AnimTime] [-z] DFiles
-l: Linear - forces linear (degree two) surfaces to be approximated
as a single polygon along their linear direction.
Although most of the time, linear direction can be exactly represented
using a single polygon, even a bilinear surface can have a free-form
shape (saddle-like) that is not representable using a single polygon.
Note that although this option will better emulate the surface shape,
it will create unnecessary polygons in cases where one is enough.
-4: Four - Generates four polygons per flat patch.
-F PolyOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4.
-o OutName: Name of output file. By default the name of the first data
file from DFiles list is used. See below on the output files.
-g: Generates the geometry file only. See below.
-T: Talkative mode. Prints processing information.
-t AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-z: Prints version number and current defaults.
Irit2Scn converts freeform surfaces and polygons into polygons in a format
that can be used by RTrace. Two files are created, one with a '.geom'
extension and one with '.scn'. Since the number of polygons can be extremely
large, the geometry is isolated in the '.geom' file and is included
(via '#include') in the main '.scn' file. The latter holds the surface
properties for all the geometry as well as viewing and RTrace specific
commands. This allows for the changing of the shading or the viewing
properties while editing small ('.scn') files.
If '-g' is specified, only the '.geom' file is created, preserving the
current '.scn' file.
In practice, it may be useful to create a low resolution approximation
of the model, change the viewing/shading parameters in the '.scn' file
until a good view and/or surface quality is found, and then run Irit2Scn once
more to create a high resolution approximation of the geometry using '-g'.
Example:
irit2scn -l -F 0 8 b58.dat
creates b58.scn and b58.geom with low resolution (FineNess of 5).
One can ray trace this scene after converting the scn file to a sff file,
using scn2sff provided with the RTrace package.
Once done with the parameter setting of RTrace, a fine approximation of the
model can be created with:
irit2scn -l -g -F 0 64 b58.dat
which will only recreate b58.geom (because of the -g option).
One can overwrite the viewing matrix by appending a new matrix
in the end of the command line, created by the display devices:
wntdrvs b58.dat
irit2scn -l -F 0 8 b58.dat irit.mat
where irit.mat is the viewing matrix created by wntdrvs. The output name,
by default, is the last input file name, so you might want to provide an
explicit name with the -o flag.
One can specify surface qualities for individual surfaces of a model.
Several such attributes are supported by Irit2Scn and can be set within
IRIT. See also the ATTRIB IRIT command.
If a certain surface should be finer/caurser than the rest of the
scene, one can set a "resolution" attribute which specifies the
relative FineNess resolution of this specific surface. Further,
"u_resolution" and "v_resolution" might be similarly used to set
relative resolution for the u or v direction only.
Example:
attrib( srf1, "resolution", 2 );
will force srf1 to have twice the default resolution, as set via the '-f'
flag.
Almost flat patches are converted to polygons. The patch can be converted
into two polygons (by subdividing along one of its diagonals) or into four
by introducing a new point at the patch center. This behavior is controlled
by the '-4' flag, but can be overwritten for individual surfaces bu setting
"twoperflat" or "fourperflat".
RTrace specific properties are controlled via the following attributes:
"SCNrefraction", "SCNtexture", "SCNsurface. Refer to the RTrace manual for
their meaning.
Example:
attrib( srf1, "SCNrefraction", 0.3 );
Surface color is controlled in two levels. If the object has an RGB
attribute, it is used. Otherwise a color as set via IRIT COLOR command
is used if set.
Example:
attrib( tankBody, "rgb", "244,164,96" );
$
IRIT2Xfg - IRIT To XFIG filter
irit2xfg [-s Size] [-t XTrans YTrans] [-I #UIso[:#VIso[:#WIso]]]
[-f PolyOpti SampTol] [-F PolyOpti FineNess] [-M] [-G] [-T]
[-a AnimTime] [-i] [-o OutName] [-z] DFiles
-s Size: Size in inches of the page. Default is 7 inches.
-t XTrans YTrans: X and Y translation. of the image. Default is (0, 0).
-I #UIso[:#VIso]: Specifies the number of isolines per surface, per
direction. If #VIso is not specified, #UIso is used for #VIso as
well.
-f PolyOpti SampTol: Controls the method used to approximate curves
into polylines. If PolyOpti == 0, equally spaced intervals are
used. For PolyOpti == 1, an adaptive subdivision that optimizes the
samples is employed. For PolyOpti == 2, SampTol (real number)
specifies the maximal allowed deviation tolerance of the piecewise
linear approximation from the original curve.
Default is 0 64 (uniform sampling with 64 samples).
-F PolygonOpti FineNess: Optimality of polygonal approximation of
surfaces. See the variable POLY_APPROX_OPT for the meaning of
FineNess. See also -4. This enforces the dump of freefrom geometry
as polygons.
-M: Dumps the control mesh/polygon as well.
-G: Dumps the freeform geometry.
-T: Talkative mode. Prints processing information.
-a AnimTime: If the data contains animation curves, evaluate and
process the scene at time AnimTime.
-i: Internal edges (created by IRIT) - default is not to
display them, and this option will force displaying them as well.
-o OutName: Name of output file. By default the name of the first data
file from DFiles list is used. See below on the output files.
-z: Prints version number and current defaults.
Irit2Xfg converts freeform surfaces and polygons into polylines in a format
that can be used by XFIG.
Example:
irit2Xfg -T -f 0 16 saddle.dat > saddle.xfg
However, one can overwrite the viewing matrix by appending a new matrix
in the end of the command line, created by the display devices:
x11drvs b58.dat
irit2Xfg -T -f 0 16 b58.dat irit.mat > saddle.xfg
where irit.mat is the viewing matrix created by x11drvs.
$
OBJ2IRIT - Wavefront OBJ format To IRIT data files
Converts Waverfront's OBJ data files into IRIT data files.
obj2irit [-m] [-o OutName] [-z] OBJFile
-m: Provides some more information on the data file(s) parsed.
-o OutName: Name of output file. By default the output goes to
stdout.
-z: Prints version number and current defaults.
obj2irit converts Wavefront's OBJ data files into IRIT data files.
The current version provides only partial support for the direct conversion
of freeform surfaces, mainly due to luck of examples of freeform surfaces
in obj format.
Example:
obj2irit -o file.dat file.obj
$
DATAFILE Format
This section describes the data file format used to exchange data between
IRIT and its accompanying tools.
[OBJECT {ATTRS} OBJNAME
[NUMBER n]
| [VECTOR x y z]
| [CTLPT POINT_TYPE {w} x y {z}]
| [STRING "a string"]
| [MATRIX m00 ... m03
m10 ... m13
m20 ... m23
m30 ... m33]
;A polyline should be drawn from first point to last. Nothing is drawn
;from last to first (in a closed polyline, last point is equal to first).
| [POLYLINE {ATTRS} #PTS ;#PTS = number of points.
[{ATTRS} x y z]
[{ATTRS} x y z]
.
.
.
[{ATTRS} x y z]
]
;Defines a closed planar region. Last point is NOT equal to first,
;and a line from last point to first should be drawn when the boundary
;of the polygon is drawn.
| [POLYGON {ATTRS} #PTS
[{ATTRS} x y z]
[{ATTRS} x y z]
.
.
.
[{ATTRS} x y z]
]
;Defines a "cloud" of points.
| [POINTLIST {ATTRS} #PTS
[{ATTRS} x y z]
[{ATTRS} x y z]
.
.
.
[{ATTRS} x y z]
]
;Defines an instance - a geometry reference (by name, SRF13 below)
;and a transformation matrix to apply to this geoemtry
| [INSTANCE SRF13
m00 ... m03
m10 ... m13
m20 ... m23
m30 ... m33]
;Defines a Bezier curve with #PTS control points. If the curve is
;rational, the rational component is introduced first.
| [CURVE BEZIER {ATTRS} #PTS POINT_TYPE
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bezier surface with #UPTS * #VPTS control points. If the
;surface is rational, the rational component is introduced first.
;Points are printed row after row (#UPTS per row), #VPTS rows.
| [SURFACE BEZIER {ATTRS} #UPTS #VPTS POINT_TYPE
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bezier triangular surface with (#PTS + 1) * #PTS / 2 control
;points, of order ORDER. If the surface is rational, the rational
;component is introduced first. Note #PTS holds number of points along
;an edge and is exactly equal to ORDER. Points are printed sequentially.
| [TRISRF BEZIER {ATTRS} #PTS POINT_TYPE
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bezier trivariate with #UPTS * #VPTS * #WPTS control
;points. If the trivariate is rational, the rational component is
;introduced first. Points are printed row after row (#UPTS per row),
;#VPTS rows, #WPTS layers (depth).
| [TRIVAR BEZIER {ATTRS} #UPTS #VPTS #WPTS POINT_TYPE
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bspline curve of order ORDER with #PTS control points. If the
;curve is rational, the rational component is introduced first.
;Note length of knot vector is equal to #PTS + ORDER.
;If curve is periodic KVP prefix the knot vector that has length of
;'Length + Order + Order - 1'.
| [CURVE BSPLINE {ATTRS} #PTS ORDER POINT_TYPE
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;Knot vector
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bspline surface with #UPTS * #VPTS control points, of order
;UORDER by VORDER. If the surface is rational, the rational component
;is introduced first.
;Points are printed row after row (#UPTS per row), #VPTS rows.
;If surface is periodic in some direction KVP prefix the knot vector
;that has length of 'Length + Order + Order - 1'.
| [SURFACE BSPLINE {ATTRS} #UPTS #VPTS UORDER VORDER POINT_TYPE
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;U Knot vector
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;V Knot vector
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bspline triangular surface with (#PTS + 1) * #PTS / 2 control
;points, of order ORDER. If the surface is rational, the rational
;component is introduced first.
;Points are printed sequentially.
| [TRISRF BSPLINE {ATTRS} #PTS ORDER POINT_TYPE
[KV {ATTRS} kv0 kv1 kv2 ...] ;Knot vector
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a Bspline trivariate with #UPTS * #VPTS * #WPTS control
;points. If the trivariate is rational, the rational component is
;introduced first. Points are printed row after row (#UPTS per row),
;#VPTS rows, #WPTS layers (depth).
;If trivariate is periodic in some direction KVP prefix the knot vector
;that has length of 'Length + Order + Order - 1'.
| [TRIVAR BSPLINE {ATTRS} #UPTS #VPTS #WPTS UORDER VORDER WORDER POINT_TYPE
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;U Knot vector
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;V Knot vector
[KV{P} {ATTRS} kv0 kv1 kv2 ...] ;W Knot vector
[{ATTRS} {w} x y z ...]
[{ATTRS} {w} x y z ...]
.
.
.
[{ATTRS} {w} x y z ...]
]
;Defines a trimmed surface. Encapsulates a surface (can be either a
;Bspline or a Bezier surface) and prescribes its trimming curves.
;There can be an arbitrary number of trimming curves (either Bezier
; or Bspline). Each trimming curve contains an arbitrary number of
;trimming curve segments, while each trimming curve segment contains
;a parameteric representation optionally followed by a Euclidean
;representation of the trimming curve segment.
| [TRIMSRF
[SURFACE ...
]
[TRIMCRV
[TRIMCRVSEG
[CURVE ...
]
]
.
.
.
[TRIMCRVSEG
[CURVE ...
]
]
]
.
.
.
[TRIMCRV
[TRIMCRVSEG
[CURVE ...
]
]
.
.
.
[TRIMCRVSEG
[CURVE ...
]
]
]
]
]
POINT_TYPE -> E1 | E2 | E3 | E4 | E5 | P1 | P2 | P3 | P4 | P5
ATTRS -> [ATTRNAME ATTRVALUE]
| [ATTRNAME]
| [ATTRNAME ATTRVALUE] ATTRS
Some notes:
* This definition for the text file is designed to minimize the
reading time and space. All information can be read without backward
or forward referencing.
* An OBJECT must never hold different geometry types or other entities.
I.e. CURVEs, SURFACEs, and POLYGONs must all be in different OBJECTs.
* Attributes should be ignored if not needed. The attribute list may have
any length and is always terminated by a token that is NOT 'verb+[+'. This
simplifies and disambiguates the parsing.
* Comments may appear between 'verb+[+OBJECT ...verb+]+' blocks, or
immediately after OBJECT OBJNAME, and only there.
A comment body can be anything not containing the 'verb+[+' or the
'verb+]+' tokens (signals start/end of block). Some of the comments in
the above definition are illegal and appear there only of the sake
of clarity.
* It is preferred that geometric attributes such as NORNALs will be saved in
the geometry structure level (POLYGON, CURVE or vertices) while graphical
and others such as COLORs will be saved in the OBJECT level.
* Objects may be contained in other objects to an arbitrary level.
Here is an example that exercises most of the data format:
This is a legal comment in a data file.
[OBJECT DEMO
[OBJECT REAL_NUM
And this is also a legal comment.
[NUMBER 4]
]
[OBJECT A_VECTOR
[VECTOR 1 2 3]
]
[OBJECT CTL_POINT
[CTLPT E3 1 2 3]
]
[OBJECT STR_OBJ
[STRING "string"]
]
[OBJECT UNIT_MAT
[MATRIX
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
]
]
[OBJECT [COLOR 4] POLY1OBJ
[POLYGON [PLANE 1 0 0 0.5] 4
[-0.5 0.5 0.5]
[-0.5 -0.5 0.5]
[-0.5 -0.5 -0.5]
[-0.5 0.5 -0.5]
]
[POLYGON [PLANE 0 -1 0 0.5] 4
[0.5 0.5 0.5]
[-0.5 0.5 0.5]
[-0.5 0.5 -0.5]
[0.5 0.5 -0.5]
]
]
[OBJECT [COLOR 63] ACURVE
[CURVE BSPLINE 16 4 E2
[KV 0 0 0 0 1 1 1 2 3 4 5 6 7 8 9 10 11 11 11 11]
[0.874 0]
[0.899333 0.0253333]
[0.924667 0.0506667]
[0.95 0.076]
[0.95 0.76]
[0.304 1.52]
[0.304 1.9]
[0.494 2.09]
[0.722 2.242]
[0.722 2.318]
[0.38 2.508]
[0.418 2.698]
[0.57 2.812]
[0.57 3.42]
[0.19 3.572]
[0 3.572]
]
]
[OBJECT [COLOR 2] SOMESRF
[SURFACE BEZIER 3 3 E3
[0 0 0]
[0.05 0.2 0.1]
[0.1 0.05 0.2]
[0.1 -0.2 0]
[0.15 0.05 0.1]
[0.2 -0.1 0.2]
[0.2 0 0]
[0.25 0.2 0.1]
[0.3 0.05 0.2]
]
]
]
$
BUGS and LIMITATIONS
Like any program of more than one line, it is far from being perfect.
Some limitations, as well as simplifications, are laid out below.
* If the intersection curve of two objects falls exactly on polygon
boundaries, for all polygons, the system will scream that the two objects
do not intersect at all. Try to move one by EPSILON into the other.
I probably should fix this one - it is supposed to be relatively easy.
* Avoid degenerate intersections that result with a point or a line.
They will probably cause wrong propagation of the inner and outer part of
one object relative to the other. Always extend your object beyond the
other object.
* If two objects have no intersection in their boundary, IRIT assumes they
are disjoint: a union simply combines them, and the other Boolean
operators return a NULL object. One should find a FAST way (3D Jordan
theorem) to find the relation between the two (A in B, B in A, A
disjoint B) and according to that, make a decision.
* Since the boolean sum implementation constructs ruled surfaces with
uniform speed, it might return a somewhat incorrect answer, given
non-uniform input curves.
* The parser is out of hand and is difficult to maintain. There are several
memory leaks there that one should fix.
* The X11 driver has no menu support (any easy way to have menus using
Xlib!?).
* IBM R6000 fails to run the drivers in -s- mode.
* Rayshade complains a lot about degenerate polygons on irit2ray output.
To alleviate the problem, change the 'equal' macro in common.h in libcommon
of rayshade from EPSILON (1e-5) to 1e-7 or even lower.
* On the motif-based drivers (xmtdrvs etc.) clicking the mouse left and
right of the scale's button produces stepped transformations. This
step size is constant, and is not proportional to the distance between
the mouse's position and the position of the button. The reason for the
flaw is incorrect callback information returned from the scale in
repeattive mode.
* Binary data files are not documented, nor will they be. They might
change in the future and are in fact machine dependend. Hence, one
platform might fail to read the other's binary data file.
$