Re: Proposal: immediate strings

[Stefan Monnier <monnier@iro.umontreal.ca>]

> That might work.  But this raises another idea: assuming most tiny strings
> don't have text properties, won't it improve performance overall if
> any string with text properties is forced to be an ordinary string, so
> that immediate strings can reuse the rarely-used 'intervals' member
> for data?

That's part of the reason why I'm interested in his statistics about
which strings have text properties.

But even if there's a potential gain here, there's still the issue that
we may then have to convert an immediate string into an ordinary string
when text-properties are added, which will imply basically relocating
the string data, with the usual risk that we might happen to do that
right in the middle of a loop keeping a pointer to the inside of the
string's data.

> Another thought that comes to mind, is that we could leave
> the mark bit where it is (the most significant bit of 'size'),
> and reserve 'size' values >= (EMACS_INT / 2 & ~0xffff)

No need to be that careful: the `size' field has 2 free bits (since the
size is limited by what integer can fit in a Lisp_Object), and only 1 is
used (for the markbit), so we could just grab a second bit from `size'
for the `immbit'.


        Stefan