Optimize term_feed scalar dispatch, branchless range search, and invalid subparts

1 files changed, 11 insertions, 10 deletions

diff --git a/include/term/unicode/width.h b/include/term/unicode/width.h
index 388ae6c..a7dfe66 100644
--- a/include/term/unicode/width.h
+++ b/include/term/unicode/width.h
@@ -11,17 +11,16 @@ struct unicode_range { uint32_t lo, hi; };
 
 static inline bool in_ranges(const struct unicode_range *ranges, size_t count, uint32_t cp)
 {
-    size_t lo = 0, hi = count;
-    while (lo < hi) {
-        size_t mid = lo + (hi - lo) / 2;
-        if (cp < ranges[mid].lo)
-            hi = mid;
-        else if (cp > ranges[mid].hi)
-            lo = mid + 1;
-        else
-            return true;
+    if (count == 0 || cp < ranges[0].lo || cp > ranges[count - 1].hi)
+        return false;
+    const struct unicode_range *base = ranges;
+    size_t n = count;
+    while (n > 1) {
+        size_t half = n / 2;
+        base = (base[half].lo <= cp) ? base + half : base;
+        n -= half;
     }
-    return false;
+    return cp <= base->hi;
 }
 
 #define IN_TABLE(table, cp) in_ranges(table, sizeof(table) / sizeof(table[0]), cp)
@@ -47,6 +46,8 @@ static inline bool unicode_emoji_presentation(uint32_t cp)
 
 static inline bool unicode_spacing_mark(uint32_t cp)
 {
+    if (cp < 0x903 || ((cp - 0xABEDu) < (0x11000u - 0xABEDu)) || cp > 0x1D281)
+        return false;
     return IN_TABLE(spacing_mark, cp);
 }