I'm excited to announce that the text metric functions that I've been
working on are finally in the development snapshot.  (OK, I'm easily
excited.)

What they do:  You give textmetrics() parameters equivalent to the
parameters to text() - a string, size, font, et cetera - and the
textmetrics() function tells you various things about how that string would
be rendered.  It tells you the bounding box of the actual object created,
how far above and below the baseline the text extends, where the next
string would be placed, et cetera.  The related fontmetrics() function
gives you some font-global measurements and, perhaps most usefully, tells
you the name of the actual font selected.

What they don't do:  They don't give you the list of points that make up
the text.  (Maybe in a future project, but maybe that would be better done
in a generic way that works for any geometric object.)

Also interesting is that this project introduces a new data type - for the
moment called an "object", but I don't know that that's really been settled
on, given that that name is also used for geometric objects.  An object is,
like an object in JavaScript or a dictionary in Python, a collection of
names and values.  Objects are  important to this project because they
allow the function to return several values in a structured and
self-documenting way.  You can retrieve values out of an object using
either "object.name" or "object[string]", you can get all of the keys for
the object, you can echo an object, you can convert an object into a
string, and you can test for whether a particular data item is an object
using is_object().  You cannot yet create your own objects.

These new features are still experimental.  To use them, you have to turn
them on at Edit / Preferences / Features.

Comments solicited.

Here's some draft documentation.

textmetrics() returns information about the size and positioning that
would be applied to text if laid out as specified by its arguments.

The arguments to textmetrics() are exactly the same as the arguments to
text().

The return value from textmetrics() is an object with these properties:

- position: an [x,y] pair giving the lower left corner of the smallest
box that would completely surround the text.
- size: an [x,y] pair giving the size of that box.
- ascent: the vertical distance (normally positive) from the baseline
of the text to the highest point in the text.
- descent: the vertical distance (normally zero or negative) from the
baseline of the text to the lowest point in the text.
- offset: an [x,y] pair giving the distance from the origin to the
starting point of the baseline of the text. This value is normally [0,0]
but can be non-zero when using non-default alignments.
- advance: an [x,y] pair giving the distance from the starting point
for this text to the starting point for a subsequent piece of text.
Informally, it says how far the pen should be moved before starting the
next piece of text. This value is directly helpful only with default
alignments - horizontal text aligned left/baseline and vertical text
aligned center/top.

Note:  as with any array, you can use an object-like ".x" or ".y" to refer
to the two entries in [x,y] pairs.  Thus you can refer to either
obj.size[0] or obj.size.x to get the X dimension of the text.

Here's a sample output (reformatted for readability):

ECHO: {
position = [2.2784, -5.7728];
size = [87.0362, 24.8832];
ascent = 19.1104;
descent = -5.7728;
offset = [0, 0];
advance = [89.5186, 0];
}

and here's an illustration showing some of the values:

fontmetrics() returns information about the specified font and size.

The arguments to fontmetrics() are the size and font name.

The return value from fontmetrics() is an object with these properties:

- nominal: an object with ascent and descent properties describing the
"usual" ascent and descent of the font. For most fonts, this will be the
ascent of upper-case letters and the descent of lower-case letters with
descenders.
- max: an object with ascent and descent properties describing the
maximum ascent and descent of the font.
- interline: gives, as a positive number, the designed inter-line
spacing for the font.
- font: an object with members family and style that gives the name
and style of the selected font. These may not be the same as the name
supplied, if the specified font is not available, or if a generic name like
"sans" or "serif" is specified, or if only a style is specified.

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org

At first glance, this looks awesome!  I'm not at a point that I can
install a dev build, but I look forward to trying this out in the
future.  

Some more examples of how these are used would be nice.

I am especially excited about your new "objects" and being able to
access a value by "obj.size.x" syntax.  Being able to define our own
"objects" would certainly raise OpenSCAD to a higher plane of
existence!  ;-)

I should mention that there are a couple of subtle incompatible changes.
These behaviors are not controlled by the experimental flag.

All inter-glyph spacing was previously shrunk by 2.4%.  Nobody remembers
why, and it sure seemed like it was a mistake.  I got rid of that, so the
text is at its designed spacing.  If you need to, I think you can
compensate by setting spacing to 1/1.024.

Non-default alignment on vertical text was just wrong.  As I recall, the
default stayed pretty much the same, but I fixed the other variations.

Note:  I think that alignment behavior on right-to-left and bottom-to-top
text is still wrong; for instance, I think that the default for RTL text
should be halign=right.  But that's not something I want to fix today.

OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org

I already use https://www.thingiverse.com/thing:3004457
https://www.thingiverse.com/thing:3004457 by Alexander Pruss to make
our garden plant signs.  Looking forward to seeing what yours does.