fast getX(.is_symbol)

src/benchmark/IsSymbol.zig

@@ -11,6 +11,7 @@ const Benchmark = @import("Benchmark.zig");
 const options = @import("options.zig");
 const UTF8Decoder = @import("../terminal/UTF8Decoder.zig");
 const symbols = @import("../unicode/symbols.zig");
+const uucode = @import("uucode");
 
 const log = std.log.scoped(.@"is-symbol-bench");
 
@@ -103,7 +104,7 @@ fn stepUucode(ptr: *anyopaque) Benchmark.Error!void {
             const cp_, const consumed = d.next(c);
             assert(consumed);
             if (cp_) |cp| {
-                std.mem.doNotOptimizeAway(symbols.isSymbol(cp));
+                std.mem.doNotOptimizeAway(uucode.getX(.is_symbol, cp));
             }
         }
     }

src/build/uucode_config.zig

@@ -90,6 +90,29 @@ const grapheme_boundary_class = config.Extension{
     },
 };
 
+fn computeIsSymbol(cp: u21, data: anytype, backing: anytype, tracking: anytype) void {
+    _ = cp;
+    _ = backing;
+    _ = tracking;
+    const block = data.block;
+    data.is_symbol = data.general_category == .other_private_use or
+        block == .dingbats or
+        block == .emoticons or
+        block == .miscellaneous_symbols or
+        block == .enclosed_alphanumerics or
+        block == .enclosed_alphanumeric_supplement or
+        block == .miscellaneous_symbols_and_pictographs or
+        block == .transport_and_map_symbols;
+}
+
+const is_symbol = config.Extension{
+    .inputs = &.{ "block", "general_category" },
+    .compute = &computeIsSymbol,
+    .fields = &.{
+        .{ .name = "is_symbol", .type = bool },
+    },
+};
+
 pub const tables = [_]config.Table{
     .{
         .extensions = &.{wcwidth},
@@ -113,4 +136,10 @@ pub const tables = [_]config.Table{
             grapheme_boundary_class.field("grapheme_boundary_class"),
         },
     },
+    .{
+        .extensions = &.{is_symbol},
+        .fields = &.{
+            is_symbol.field("is_symbol"),
+        },
+    },
 };

src/unicode/symbols.zig

@@ -31,15 +31,8 @@ pub const table = table: {
 /// In the future it may be prudent to expand this to encompass more
 /// symbol-like characters, and/or exclude some PUA sections.
 pub fn isSymbol(cp: u21) bool {
-    const block = uucode.get(.block, cp);
-    return uucode.get(.general_category, cp) == .other_private_use or
-        block == .dingbats or
-        block == .emoticons or
-        block == .miscellaneous_symbols or
-        block == .enclosed_alphanumerics or
-        block == .enclosed_alphanumeric_supplement or
-        block == .miscellaneous_symbols_and_pictographs or
-        block == .transport_and_map_symbols;
+    // TODO: probably can remove this method and just call uucode directly
+    return uucode.getX(.is_symbol, cp);
 }
 
 /// Runnable binary to generate the lookup tables and output to stdout.