2025-08-23 Sat 23:13 Piece table

How is it that now after spending so much time on my text editor I'm
finding all those resources that would be very helpful and interesting
back when I started.  Few years ago when I was actively researching
the topic it was impossible to find anything beyond few small projects
with line based editing.

But now everything is screaming at me that gap buffer might not be the
right choice.  I just learned about the Piece table [1] which at first
sounded awful but after longer consideration I admit that it has more
than one nice property.  One of which is that undo history is almost
for free.  Then there is this blog post [2] linking to yet another
resources which talk about VS code approach to text buffer [3] and
criticism of gap buffer [4] which apparently is no good for multi
cursor edits.

Well I have not got to the point of implementing history and multiple
cursors in my text editor development journey.  It might just be that
macro system that I will make will made multiple cursors obsolete and
most probably the advantage of easy undo history in piece table will
not work with files as big as few GB which my editor suppose to
handle.  We will see.

What strikes me the most is that piece table looks almost exactly like
undo history session structure I designed with my brother during
random discussion on the topic.  It's not the first time when someone
reinvents something.  Looks like given certain design decisions,
limitations and constrains the solution presents itself.

> I find, that if I just sit down to think...
> The solution presents itself!
> - Professor Henry Jones

[1] https://en.wikipedia.org/wiki/Piece_table
[2] https://cdacamar.github.io/data%20structures/algorithms/benchmarking/text%20editors/c++/editor-data-structures/
[3] https://code.visualstudio.com/blogs/2018/03/23/text-buffer-reimplementation
[4] https://nullprogram.com/blog/2017/09/07/

EOF