The ceiling in change_font_size is ten times the configured font_size, so
the largest size you can reach depends on the size you configured. With
font_size 6 zooming stops at 60pt, while font_size 20 reaches 200pt. That
is backwards: a small configured font is if anything a reason to want a
larger temporary zoom, not a smaller one.
Use the larger of that multiple and an absolute 256pt. No existing setup
loses range, since the multiple still wins whenever it is the bigger of
the two. The bounds and the clamp move into options/utils.py so the
numbers are in one place and can be tested directly.
increase_font_size also had its own cap of five times font_size. That was
the old limit, missed when 25d8d9fe6 raised change_font_size to ten, and
because the wrapper clamps before delegating, the stale bound is the one
that wins. Both legacy wrappers can simply drop their bounds: they go
through change_font_size, which already clamps.
This is actually not a security boundary for this protocol. Since it is
explicitly *not* designed to be proof against malicious clients.
However, to avoid more AI generated security reports against it...
Alpha blending of graphics protocol images and animation frames is
2-3.5x faster and glyph alpha masks are composited onto canvases using
the same vectorized primitives
This is not strictly necessary as their callers already validate bounds,
however, I have now got two separate false security reports for this so
rather than waste my time responding to these, just add the bounds
checking.
Pre-edit text was drawn in reverse video, which is a solid inverted block
moving through the line as you compose. Draw it as italic with a dashed
underline in the highlight color instead.
Underline rather than reverse video is what other terminals do: VTE renders the
pre-edit run with the input method's own attributes and marks only the cursored
cell distinctly, and foot underlines the run explicitly, excluding the part
under the cursor. Dashed avoids the underline styles applications already use:
curly for spell checking and straight for hyperlinks, which is the collision
that made reverse video look like the safer choice.
The styling is applied to the overlay's own cursor and restored afterwards,
alongside the wrap, cursor-visibility and insert-replace modes already saved
there. The previous code toggled reverse with XOR, which restored itself; the
three new attributes are saved and put back explicitly.
Closes#10382
Optimise the AVX 2 implementation reducing number of operations per block
and also add an AVX-512 implementation used when the CPU supports it at
runtime.
Mostly comes from detecting the case of runs of ASCII printable
characters and using a special fast path for them. Improves benchmarks
as per benchmark.py by 15-30% for text heavy workloads