Re: Allow scripts to be defined either as glyphs or stencils. (issue 348120043 by address@hidden)

[dak]

On 2019/02/22 21:51:39, Valentin Villenave wrote:
> On 2019/02/22 21:00:08, dak wrote:
> > You are confusing stencils with stencil expressions.  Stencils
satisfy
> > ly:stencil? and you can extract their stencil expression (which
usually is a
> > pair) with ly:stencil-expr .

> Yes, but what we’re dealing with in this function is a stencil
expression, isn’t
> it? (At least, as long as it isn’t unsmobbed.)

There is some fundamental confusion here.  A stencil expression is a
list.  It never satisfies the ly:stencil? predicate and always satisfies
the pair? predicate.  Calling unsmob<Stencil> on its SCM representation
will deliver a 0 pointer.  It is not a smob.  A Stencil is a smob.  It
will satisfy the ly:stencil? predicate and never will satisfy the pair?
predicate.  Calling unsmob<Stencil> on its SCM representation will
deliver a valid Stencil * .  Calling ly:stencil-expr on it will deliver
its stencil expression.  A Stencil type contains both a stencil
expression and a bounding box.  Those are basically indendent components
of a Stencil .

> If I do

> if (scm_is_pair (s) {} else *unsmob<Stencil> (s)

> the stencil expression in "s" will never reach the "else" subroutine.

Wrong.

> > I don't understand where the difference in what you do with
script-stencil as
> a
> > property and what you would do with stencil as a property would lie.

> Forgive me for not understanding what you mean right away.  Right now,
> Script.stencil is defined as ly:script-interface::print which is where
glyphs
> are looked up (or, with this new patch, stencil expressions), then
printed.  Are
> you suggesting that we scrap this interface and make Script.stencil
point to a
> Scheme callback instead?

ly:script-interface::print is a Scheme callback (it may or may not be
implemented in C++ but that's not really relevant).  What does
overriding script-stencil buy you that overriding stencil would not
equally well do?  I want to understand what this achieves in a better
way than what we have already.

https://codereview.appspot.com/348120043/