(unicode-category-table): Setup unicode-category-table.

1 files changed, 20 insertions, 0 deletions

diff --git a/lisp/international/characters.el b/lisp/international/characters.el
index 6e8cddefe57..5543650ff7c 100644
--- a/lisp/international/characters.el
+++ b/lisp/international/characters.el
@@ -1116,6 +1116,26 @@ Setup char-width-table appropriate for non-CJK language environment."
  'tibetan)
 
 
+;;; Setting unicode-category-table.
+
+;; This macro is to build unicode-category-table at compile time so
+;; that C code can access the table efficiently.
+(defmacro build-unicode-category-table ()
+  (let ((table (make-char-table 'unicode-category-table nil)))
+    (dotimes (i #x110000)
+      (if (or (< i #xD800)
+              (and (> i #xF900) (< i #x30000))
+              (and (> i #xE0000) (< i #xE0200)))
+          (aset table i (get-char-code-property i 'general-category))))
+    (set-char-table-range table '(#xE000 . #xF8FF) 'Co)
+    (set-char-table-range table '(#xF0000 . #xFFFFD) 'Co)
+    (set-char-table-range table '(#x100000 . #x10FFFD) 'Co)
+    (optimize-char-table table 'eq)
+    table))
+
+(setq unicode-category-table (build-unicode-category-table))
+
+
 ;;; Setting word boundary.
 
 (defun next-word-boundary-han (pos limit)