Mercurial > octave-nkf
view test/bug-31371.tst @ 20308:b8c1b3e9aa91 stable
Fix segfault when '[]' used for row_vector_property.
* graphics.in.h (row_vector_property::add_constraint (octave_idx_type len)):
Remove validation that accepts 0x0 matrix.
* graphics.in.h (image::init): Call add_constraint() to allow 0x0 since
xdata, ydata can legitimately be '[]' for these properties.
author | Rik <rik@octave.org> |
---|---|
date | Tue, 23 Jun 2015 09:59:54 -0700 |
parents | 6fe6ac8bbfdb |
children |
line wrap: on
line source
%!test %! % Work around MATLAB bug where f(x)(y) is invalid syntax %! % (This bug does not apply to Octave) %! %! C = @(fcn,x) fcn(x); %! C2 = @(fcn,x,y) fcn(x,y); %! %! % Church Booleans %! T = @(t,f) t; %! F = @(t,f) f; %! %! % Church Numerals %! Zero = @(fcn,x) x; %! One = @(fcn,x) fcn(x); %! Two = @(fcn,x) fcn(fcn(x)); %! Three = @(fcn,x) fcn(fcn(fcn(x))); %! Four = @(fcn,x) fcn(fcn(fcn(fcn(x)))); %! %! % Arithmetic Operations %! Inc = @(a) @(f,x) f(a(f,x)); % Increment %! Add = @(a,b) @(f,x) a(f,b(f,x)); %! Mult = @(a,b) @(f,x) a(@(x) b(f,x),x); %! Dec = @(a) @(f,x) C(a(@(g) @(h) h(g(f)), @(u) x), @(u) u); % Decrement %! Sub = @(a,b) b(Dec, a); %! %! % Renderer - Convert church numeral to "real" number %! Render = @(n) n(@(n) n+1,0); %! %! % Predicates %! Iszero = @(n) n(@(x) F, T); %! %! % Y combinator implements recursion %! Ycomb = @(f) C(@(g) f(@(x) C(g(g), x)), ... %! @(g) f(@(x) C(g(g), x))); %! %! Factorial = Ycomb(@(f) @(n) C(C2(Iszero(n), ... %! @(d) One, @(d) Mult(n, f(Dec(n)))),0)); %! %! assert (Render (Factorial (Two)), 2) %! assert (Render (Factorial (Three)), 6) %! assert (Render (Factorial (Four)), 24)