Re: [PATCH] add build-time option to use unicode9 character widths

[Peter Stephenson <p.stephenson@samsung.com>]

On Tue, 29 Nov 2016 11:53:51 -0800
Bart Schaefer <schaefer@brasslantern.com> wrote:
> On Nov 29,  5:21pm, Peter Stephenson wrote:
> } Subject: Re: [PATCH] add build-time option to use unicode9 character width
> }
> } On Tue, 29 Nov 2016 00:42:27 -0700
> } Joshua Rubin <joshua@rubixconsulting.com> wrote:
> } > This patch adds support for unicode9 character widths.
> } 
> } Thanks, that looks unproblematic as it's specially enabled --- there
> } should be some other people who can try it out now it's in.
> 
> This is almost exactly what I was suggesting we should put in a module,
> except with all those large tables loaded from a run-time external file
> of some sort.

There's a way forward, anyway.

It's separately enabled for now which gives people who want it the
ability to use it.

If it migrates to a module, --enable-unicode9 can then force the module
to be linked and loaded at boot.  Loading from an external file is then
a separate job.  There currently isn't an obvious hook to how this gets
linked in from a module, and as Sebastian has discovered the character
analysis functions really need to be lightweight (though character width
isn't among the most performance sensitive, I wouldn't think).  Autoload
seems pointless for something like this.

pws

diff --git a/NEWS b/NEWS
index 433703a..7ac144b 100644
--- a/NEWS
+++ b/NEWS
@@ -7,6 +7,12 @@ Note also the list of incompatibilities in the README file.
 Changes from 5.2 to 5.3
 -----------------------
 
+It is possible to enable character width support for Unicode 9 by
+congiuring with `--enable-unicode9'; this compiles in some additional
+tables.  At some point this support may move into a module, in which
+case the configure option will be changed to cause the module to be
+permanently loaded.
+
 The new word modifier ':P' computes the physical path of the argument.
 It is different from the existing ':a' modifier which always resolves
 '/before/here/../after' to '/before/after', and differs from the