25 KiB
Documenting OpenSCAD Code
Documentation comment blocks are all based around a single simple syntax:
// Block Name(Metadata): TitleText
// Body line 1
// Body line 2
// Body line 3
- The Block Name is one or two words, both starting with a capital letter.
- The Metadata is in parentheses. It is optional, and can contain fairly arbitrary text, as long as it doesn't include newlines or parentheses. If the Metadata part is not given, the parentheses are optional.
- A colon
:
will always follow after the Block Name and optional Metadata. - The TitleText will be preceded by a space
- The body will contain zero or more lines of text indented by three spaces after the comment markers. Each line can contain arbitrary text.
So, for example, a Figure block to show a 640x480 animated GIF of a spinning shape may look like:
// Figure(Spin,Size=640x480,VPD=444): A Cube and Cylinder.
// cube(80, center=true);
// cylinder(h=100,d=60,center=true);
Various block types don't need all of those parts, so they may look simpler:
// Topics: Mask, Cylindrical, Attachable
Or:
// Description:
// This is a description.
// It can be multiple lines in length.
Or:
// Usage: Typical Usage
// x = foo(a, b, c);
// x = foo([a, b, c, ...]);
Comments blocks that don't start with a known block header are ignored and not added to output documentation. This lets you have normal comments in your code that are not used for documentation. If you must start a comment block with one of the known headers, then adding a single extra /
or space after the comment marker, will make it be treated as a regular comment:
/// File: Foobar.scad
Block Headers
File/LibFile Blocks
All files must have either a // File:
block or a // LibFile:
block at the start. This is the place to put in the canonical filename, and a description of what the file is for. These blocks can be used interchangably, but you can only have one per file. // File:
or // LibFile:
blocks can be followed by a multiple line body that are added as markdown text after the header:
// LibFile: foo.scad
// You can have several lines of markdown formatted text here.
// You just need to make sure that each line is indented, with
// at least three spaces after the comment marker. You can
// denote a paragraph break with a comment line with three
// trailing spaces, or just a period.
// .
// The end of the block is denoted by a line without a comment.
Which outputs Markdown code that renders like:
LibFile: foo.scad
You can have several lines of markdown formatted text here. You just need to make sure that each line is indented, with at least three spaces after the comment marker. You can denote a paragraph break with a comment line with three trailing spaces, or just a period.
The end of the block is denoted by a line without a comment.
Or:
// File: Foobar.scad
// This file contains a collection of metasyntactical nonsense.
Which outputs Markdown code that renders like:
File: Foobar.scad
This file contains a collection of metasyntactical nonsense.
Includes Block
To declare what code the user needs to add to their code to include or use this library file, you can use the // Includes:
block. You should put this right after the // File:
or // LibFile:
block. This code block will also be prepended to all Example and Figure code blocks before they are evaluated:
// Includes:
// include <BOSL2/std.scad>
// include <BOSL2/beziers.scad>
Which outputs Markdown code that renders like:
Includes:
To use, add the following lines to the beginning of your file:
include <BOSL2/std.scad> include <BOSL2/beziers.scad>
CommonCode Block
If you have a block of code you plan to use throughout the file's Figure or Example blocks, and you don't actually want it displayed, you can use a // CommonCode:
block like thus:
// CommonCode:
// module text3d(text, h=0.01, size=3) {
// linear_extrude(height=h, convexity=10) {
// text(text=text, size=size, valign="center", halign="center");
// }
// }
This doesn't have immediately visible markdown output, but you can use that code in later examples:
// Example:
// text3d("Foobar");
Section Block
Section blocks take a title, and an optional body that will be shown as the description of the Section. If a body line if just a .
(dot, period), then that line is treated as a blank line in the output:
// Section: Foobar
// You can have several lines of markdown formatted text here.
// You just need to make sure that each line is indented, with
// at least three spaces after the comment marker. You can
// denote a paragraph break with a comment line with three
// trailing spaces, or just a period.
// .
// The end of the block is denoted by a line without a comment.
// or a line that is unindented after the comment.
Which outputs Markdown code that renders like:
Section: Foobar
You can have several lines of markdown formatted text here. You just need to make sure that each line is indented, with at least three spaces after the comment marker. You can denote a paragraph break with a comment line with three trailing spaces, or just a period.
The end of the block is denoted by a line without a comment. or a line that is unindented after the comment.
Sections can also include Figures; images generated from code that is not shown in a code block.
Item Blocks
Item blocks headers come in four varieties: Constant
, Function
, Module
, and Function&Module
.
The Constant
header is used to document a code constant. It should have a Description sub-block, and Example sub-blocks are recommended:
// Constant: PHI
// Description: The golden ratio phi.
PHI = (1+sqrt(5))/2;
Which outputs Markdown code that renders like:
Constant: PHI
Description: The golden ration phi.
The Module
header is used to document a module. It should have a Description sub-block. It is recommended to also have Usage, Arguments, and Example/Examples sub-blocks:
// Module: cross()
// Usage:
// cross(size);
// Description:
// Creates a 2D cross/plus shape.
// Arguments:
// size = The scalar size of the cross.
// Example(2D):
// cross(size=100);
module cross(size=1) {
square([size, size/3], center=true);
square([size/3, size], center=true);
}
Which outputs Markdown code that renders like:
Module: cross()
Usage:
- cross(size);
Description: Creates a 2D cross/plus shape.
Arguments:
Positional Arg What it does size The scalar size of the cross. Example:
cross(size=100);
GENERATED IMAGE GOES HERE
The Function
header is used to document a function. It should have a Description sub-block. It is recommended to also have Usage, Arguments, and Example/Examples sub-blocks. By default, Examples will not generate images for function blocks:
// Function: vector_angle()
// Usage:
// ang = vector_angle(v1, v2);
// Description:
// Calculates the angle between two vectors in degrees.
// Arguments:
// v1 = The first vector.
// v2 = The second vector.
// Example:
// v1 = [1,1,0];
// v2 = [1,0,0];
// angle = vector_angle(v1, v2);
// // Returns: 45
function vector_angle(v1,v2) =
acos(max(-1,min(1,(vecs[0]*vecs[1])/(norm0*norm1))));
Which outputs Markdown code that renders like:
Function: vector_angle()
Usage:
- ang = vector_angle(v1, v2);
Description: Calculates the angle between two vectors in degrees.
Arguments:
Positional Arg What it does v1
The first vector. v2
The second vector. Example:
v1 = [1,1,0]; v2 = [1,0,0]; angle = vector_angle(v1, v2); // Returns: 45
The Function&Module
header is used to document a function which has a related module of the same name. It should have a Description sub-block. It is recommended to also have Usage, Arguments, and Example/Examples sub-blocks. You should have Usage blocks for both calling as a function, and calling as a
module:
// Function&Module: oval()
// Topics: 2D Shapes, Geometry
// Usage: As a Module
// oval(rx,ry);
// Usage: As a Function
// path = oval(rx,ry);
// Description:
// When called as a function, returns the perimeter path of the oval.
// When called as a module, creates a 2D oval shape.
// Arguments:
// rx = X axis radius.
// ry = Y axis radius.
// Example(2D): Called as a Function
// path = oval(100,60);
// polygon(path);
// Example(2D): Called as a Module
// oval(80,60);
module oval(rx,ry) {
polygon(oval(rx,ry));
}
function oval(rx,ry) =
[for (a=[360:-360/$fn:0.0001]) [rx*cos(a),ry*sin(a)];
Which outputs Markdown code that renders like:
Function&Module: oval()
Topics: 2D Shapes, Geometry
Usage: As a Module
- oval(rx,ry);
Usage: As a Function
- path = oval(rx,ry);
Description: When called as a function, returns the perimeter path of the oval. When called as a module, creates a 2D oval shape.
Arguments:
Positional Arg What it does rx X axis radius. ry Y axis radius. Example: Called as a Function
path = oval(100,60); polygon(path);
GENERATED IMAGE SHOWN HERE
Example: Called as a Module
oval(80,60);
GENERATED IMAGE SHOWN HERE
These Type blocks can have a number of sub-blocks. Most sub-blocks are optional, The available standard sub-blocks are:
// Status: DEPRECATED
// Topics: Comma, Delimited, Topic, List
// Usage:
// Description:
// Arguments:
// Figure:
or// Figures
// Example:
or// Examples:
Status Block
The Status block is used to mark a function, module, or constant as deprecated:
// Status: DEPRECATED, use foo() instead
Which outputs Markdown code that renders like:
Status: DEPRECATED, use foo() instead
Topics Block
The Topics block can associate various topics with the current function or module. This can be used to make an index of Topics:
// Topics: 2D Shapes, Geometry, Masks
Which outputs Markdown code that renders like:
Topics: 2D Shapes, Geometry, Masks
Usage Block
The Usage block describes the various ways that the current function or module can be called, with the names of the arguments. By convention, the first few arguments that can be called positionally just have their name shown. The remaining arguments that should be passed by name, will have the name followed by an =
(equal sign). Arguments that are optional in the given Usage context are shown in <
and >
angle brackets:
// Usage: As a Module
// oval(rx, ry, <spin=>);
// Usage: As a Function
// path = oval(rx, ry, <spin=>);
Which outputs Markdown code that renders like:
Usage: As a Module
- oval(rx, ry, <spin=>);
Usage: As a Function
- path = oval(rx, ry, <spin=>);
Description Block
The Description block just describes the currect function, module, or constant:
// Descripton: This is the description for this function or module.
// It can be multiple lines long. Markdown syntax code will be used
// verbatim in the output markdown file, with the exception of `_`,
// which will traslate to `\_`, so that underscores in function/module
// names don't get butchered.
Which outputs Markdown code that renders like:
Description: It can be multiple lines long. Markdown syntax code will be used verbatim in the output markdown file, with the exception of
_
, which will traslate to\_
, so that underscores in function/module names don't get butchered.
Arguments Block
The Arguments block creates a table that describes the positional arguments for a function or module, and optionally a second table that describes named arguments:
// Arguments:
// v1 = This supplies the first vector.
// v2 = This supplies the second vector.
// ---
// fast = Use fast, but less comprehensive calculation method.
// dflt = Default value.
Which outputs Markdown code that renders like:
Arguments:
Positional Arg What it Does v1
This supplies the first vector. v2
The supplies the second vector.
Named Arg What it Does fast
If true, use fast, but less accurate calculation method. dflt
Default value.
Figure Block
A Figure block generates and shows an image from a script in the multi-line body, by running it in OpenSCAD. A Figures block (plural) does the same, but treats each line of the body as a separate Figure block:
// Figure: Figure description
// cylinder(h=100, d1=75, d2=50);
// up(100) cylinder(h=100, d1=50, d2=75);
// Figure(Spin,VPD=444): Animated figure that spins to show all faces.
// cube([10,100,50], center=true);
// cube([100,10,30], center=true);
// Figures:
// cube(100);
// cylinder(h=100,d=50);
// sphere(d=100);
Which outputs Markdown code that renders like:
Figure 1: Figure description GENERATED IMAGE SHOWN HERE
Figure 2: Animated figure that spins to show all faces. GENERATED IMAGE SHOWN HERE
Figure 3: GENERATED IMAGE OF CUBE SHOWN HERE
Figure 4: GENERATED IMAGE OF CYLINDER SHOWN HERE
Figure 5: GENERATED IMAGE OF SPHERE SHOWN HERE
The metadata of the Figure block can contain various directives to alter how the image will be generated. These can be comma separated to give multiple metadata directives:
NORENDER
: Don't generate an image for this example, but show the example text.Hide
: Generate, but don't show script or image. This can be used to generate images to be manually displayed in markdown text blocks.2D
: Orient camera in a top-down view for showing 2D objects.3D
: Orient camera in an oblique view for showing 3D objects.VPD=440
: Force viewpoint distance$vpd
to 440.VPT=[10,20,30]
Force the viewpoint translation$vpt
to[10,20,30]
.VPR=[55,0,600]
Force the viewpoint rotation$vpr
to[55,0,60]
.Spin
: Animate camera orbit around the[0,1,1]
axis to display all sides of an object.FlatSpin
: Animate camera orbit around the Z axis, above the XY plane.Anim
: Make an animation where$t
varies from0.0
to almost1.0
.FrameMS=250
: Sets the number of milliseconds per frame for spins and animation.Small
: Make the image small sized.Med
: Make the image medium sized.Big
: Make the image big sized.Huge
: Make the image huge sized.Size=880x640
: Make the image 880 by 640 pixels in size.Render
: Force full rendering from OpenSCAD, instead of the normal preview.Edges
: Highlight face edges.NoAxes
: Hides the axes and scales.
Example Block
An Example block shows a script, and possibly shows an image generated from it.
The script is in the multi-line body. The Examples
(plural) block does
the same, but it treats eash body line as a separate Example bloc to show.
Any images, if generated, will be created by running it in OpenSCAD:
// Example: Example description
// cylinder(h=100, d1=75, d2=50);
// up(100) cylinder(h=100, d1=50, d2=75);
// Example(Spin,VPD=444): Animated shape that spins to show all faces.
// cube([10,100,50], center=true);
// cube([100,10,30], center=true);
// Examples:
// cube(100);
// cylinder(h=100,d=50);
// sphere(d=100);
Which outputs Markdown code that renders like:
Example 1: Example description
cylinder(h=100, d1=75, d2=50); up(100) cylinder(h=100, d1=50, d2=75);
GENERATED IMAGE SHOWN HERE
Example 2: Animated shape that spins to show all faces.
cube([10,100,50], center=true); cube([100,10,30], center=true);
GENERATED IMAGE SHOWN HERE
Example 3:
cube(100);
GENERATED IMAGE OF CUBE SHOWN HERE
Example 4:
cylinder(h=100,d=50);
GENERATED IMAGE OF CYLINDER SHOWN HERE
Example 5:
sphere(d=100);
GENERATED IMAGE OF SPHERE SHOWN HERE
The metadata of the Example block can contain various directives to alter how the image will be generated. These can be comma separated to give multiple metadata directives:
NORENDER
: Don't generate an image for this example, but show the example text.Hide
: Generate, but don't show script or image. This can be used to generate images to be manually displayed in markdown text blocks.2D
: Orient camera in a top-down view for showing 2D objects.3D
: Orient camera in an oblique view for showing 3D objects. Often used to force an Example sub-block to generate an image in Function and Constant blocks.VPD=440
: Force viewpoint distance$vpd
to 440.VPT=[10,20,30]
Force the viewpoint translation$vpt
to[10,20,30]
.VPR=[55,0,600]
Force the viewpoint rotation$vpr
to[55,0,60]
.Spin
: Animate camera orbit around the[0,1,1]
axis to display all sides of an object.FlatSpin
: Animate camera orbit around the Z axis, above the XY plane.Anim
: Make an animation where$t
varies from0.0
to almost1.0
.FrameMS=250
: Sets the number of milliseconds per frame for spins and animation.Small
: Make the image small sized.Med
: Make the image medium sized.Big
: Make the image big sized.Huge
: Make the image huge sized.Size=880x640
: Make the image 880 by 640 pixels in size.Render
: Force full rendering from OpenSCAD, instead of the normal preview.Edges
: Highlight face edges.NoAxes
: Hides the axes and scales.
Modules will default to generating and displaying the image as if the 3D
directive is given. Functions and constants will default to not generating
an image unless 3D
, Spin
, FlatSpin
or Anim
is explicitly given.
If any lines of the Example script begin with --
, then they are not shown in
the example script output to the documentation, but they are included in the
script used to generate the example image, without the --
, of course:
// Example: Multi-line example.
// --$fn = 72; // Lines starting with -- aren't shown in docs example text.
// lst = [
// "multi-line examples",
// "are shown in one block",
// "with a single image.",
// ];
// foo(lst, 23, "blah");
Creating Custom Block Headers
If you have need of a non-standard documentation block in your docs, you can declare the new block type using DefineHeader:
. This has the syntax:
// DefineHeader(TYPE): NEWBLOCKNAME
Where NEWBLOCKNAME is the name of the new block header, and TYPE defines the behavior of the new block. TYPE can be one of:
Generic
: Show both the TitleText and body.Text
: Show the TitleText as the first line of the body.Label
: Show only the TitleText and no body.NumList
: Shows TitleText, and the body lines in a numbered list.BulletListList
: Shows TitleText, and the body lines in a bullet list.Table
: Shows TitleText, and body lines in a definition table.Figure
: Shows TitleText, and an image rendered from the script in the Body.Example
: Like Figure, but also shows the body as an example script.
Generic Block Type
The Generic block header type takes both title and body lines and generates a markdown block that has the block header, title, and a following body:
// DefineHeader(Generic): Result
// Result: For Typical Cases
// Does typical things.
// Or something like that.
// Result: For Atypical Cases
// Performs an atypical thing.
Which outputs Markdown code that renders like:
Result: For Typical Cases
Does typical things. Or something like that.
Result: For Atypical Cases
Performs an atypical thing.
Text Block Type
The Text block header type is similar to the Generic type, except it merges the title into the body. This is useful for allowing single-line or multi- line blocks:
// DefineHeader(Text): Reason
// Reason: This is a simple reason.
// Reason: This is a complex reason.
// It is a multi-line explanation
// about why this does what it does.
Which outputs Markdown code that renders like:
Reason:
This is a simple reason.
Reason:
This is a complex reason. It is a multi-line explanation about why this does what it does.
Label Block Type
The Label block header type takes just the title, and shows it with the header:
// DefineHeader(Label): Regions
// Regions: Antarctica, New Zealand
// Regions: Europe, Australia
Which outputs Markdown code that renders like:
Regions: Antarctica, New Zealand Regions: Europe, Australia
NumList Block Type
The NumList block header type takes both title and body lines, and outputs a numbered list block:
// DefineHeader(NumList): Steps
// Steps: How to handle being on fire.
// Stop running around and panicing.
// Drop to the ground.
// Roll on the ground to smother the flames.
Which outputs Markdown code that renders like:
Steps: How to handle being on fire.
- Stop running around and panicing.
- Drop to the ground.
- Roll on the ground to smother the flames.
BulletList Block Type
The BulletList block header type takes both title and body lines:
// DefineHeader(BulletList): Side Effects
// Side Effects: For Typical Uses
// The variable `foo` gets set.
// The default for subsequent calls is updated.
Which outputs Markdown code that renders like:
Side Effects: For Typical Uses
- The variable $foo gets set.
- The default for subsequent calls is updated.
Table Block Type
The Table block header type outputs a header block with the title, followed by
one or more tables. This is genertally meant for definition lists. The header
names are given in the DefineHeader metadata. Header names are separated by
|
(vertical bar, or pipe) characters, and sets of headers (for multiple
tables) are separated by ||
(two vertical bars). A header that starts with
the ^
(hat, or circumflex) character, will cause the items in that column
to be surrounded by `foo` literal markers. Cells in the body content are
separated by =
(equals signs):
// DefineHeader(Table:^Link Name|Description): Anchors
// Anchors: by Name
// "link1" = Anchor for the joiner Located at the back side of the shape.
// "a"/"b" = Anchor for the joiner Located at the front side of the shape.
Which outputs Markdown code that renders like:
Anchors: by Name
Link Name Description "link1"
Anchor for the joiner at the back side of the shape. "a"
/"b"
Anchor for the joiner at the front side of the shape.
You can have multiple subtables, separated by a line with only three dashes: ---
:
// DefineHeader(Table:^Pos Arg|What it Does||^Names Arg|What it Does): Args
// Args:
// foo = The foo argument.
// bar = The bar argument.
// ---
// baz = The baz argument.
// qux = The baz argument.
Which outputs Markdown code that renders like:
Args:
Pos Arg What it Does foo
The foo argument. bar
The bar argument.
Named Arg What it Does baz
The baz argument. qux
The qux argument.
Defaults Configuration
The openscad_decsgen
script looks for an .openscad_docsgen_rc
file in
the source code directory it is run in. In that file, you can give a few
defaults for what files will be processed, and where to save the generated
markdown documentation.
To ignore specific files, to prevent generating documentation for them, you can use the IgnoreFiles block. Note that the commentline prefix is not needed in the configuration file:
IgnoreFiles:
ignored1.scad
ignored2.scad
To prioritize the ordering of files when generating the Table of Contents and other indices, you can use the PrioritizeFiles block:
PrioritizeFiles:
file1.scad
file2.scad
To specify what directory to write the markdown output documentation to, you can use the DocsDirectory block:
DocsDirectory: wiki_dir
You can also use the DefineHeader block in the config file to make custom block headers:
DefineHeader(Text): Returns
DefineHeader(BulletList): Side Effects
DefineHeader(Table:^Anchor Name|Position): Extra Anchors