6703 - new types: code-point and grapheme

Both have the same size: 4 bytes.

So far I've just renamed print-byte to print-grapheme, but it still behaves
the same.

I'm going to support printing code-points next, but grapheme 'clusters'
spanning multiple code-points won't be supported for some time.

304screen.subx

@@ -131,7 +131,8 @@ $print-string-to-real-screen:end:
     5d/pop-to-ebp
     c3/return
 
-print-byte-to-real-screen:  # c: byte
+# currently only supports ascii
+print-grapheme-to-real-screen:  # c: grapheme
     # . prologue
     55/push-ebp
     89/<- %ebp 4/r32/esp
@@ -142,7 +143,7 @@ print-byte-to-real-screen:  # c: byte
     68/push 4/imm32/size
     89/<- %ecx 4/r32/esp
     (write 2 %ecx)
-$print-byte-to-real-screen:end:
+$print-grapheme-to-real-screen:end:
     # . reclaim locals
     81 0/subop/add %esp 8/imm32
     # . restore registers

400.mu

@@ -142,7 +142,7 @@ sig real-screen-size -> nrows/eax: int, ncols/ecx: int
 sig clear-real-screen
 sig move-cursor-on-real-screen row: int, column: int
 sig print-string-to-real-screen s: (addr array byte)
-sig print-byte-to-real-screen c: byte
+sig print-grapheme-to-real-screen c: grapheme
 sig print-int32-hex-to-real-screen n: int
 sig reset-formatting-on-real-screen
 sig start-color-on-real-screen fg: int, bg: int

404screen.mu

@@ -82,6 +82,7 @@ $clear-screen:body: {
   {
     break-if-=
     # fake screen
+    var space/edi: grapheme <- copy 0x20
     move-cursor screen, 1, 1
     var screen-addr/esi: (addr screen) <- copy screen
     var i/eax: int <- copy 1
@@ -94,7 +95,7 @@ $clear-screen:body: {
       {
         compare j, *ncols
         break-if->
-        print-byte screen, 0x20  # space
+        print-grapheme screen, space
         j <- increment
         loop
       }
@@ -136,13 +137,13 @@ $print-string:body: {
 }
 }
 
-fn print-byte screen: (addr screen), c: byte {
-$print-byte:body: {
+fn print-grapheme screen: (addr screen), c: grapheme {
+$print-grapheme:body: {
   compare screen, 0
   {
     break-if-!=
-    print-byte-to-real-screen c
-    break $print-byte:body
+    print-grapheme-to-real-screen c
+    break $print-grapheme:body
   }
   {
     break-if-=

apps/browse.mu

@@ -96,7 +96,8 @@ $line-loop: {
       update-attributes c, r
       compare c, 0xa  # newline
       break-if-=  # no need to print newlines
-      print-byte 0, c
+      var g/eax: grapheme <- copy c
+      print-grapheme 0, g
       col <- increment
       loop
     }
@@ -176,6 +177,7 @@ fn dump in: (addr buffered-file) {
   var c/eax: byte <- read-byte-buffered in
   compare c, 0xffffffff  # EOF marker
   break-if-=
-  print-byte 0, c
+  var g/eax: grapheme <- copy c
+  print-grapheme 0, g
   loop
 }

apps/mu.subx

@@ -412,8 +412,10 @@ Type-id:  # (stream (addr array byte))
   # some SubX types deliberately left undefined in Mu; they can only be operated on using SubX primitives
   "stream"/imm32  # 11
   "slice"/imm32  # 12
+  "code-point"/imm32  # 13; smallest scannable unit from a text stream
+  "grapheme"/imm32  # 14; smallest printable unit; will eventually be composed of multiple code-points, but currently corresponds 1:1
   # Keep Primitive-type-ids in sync if you add types here.
-                                          0/imm32 0/imm32 0/imm32
+                                                          0/imm32
   0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32
   0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32
   0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32 0/imm32

apps/print-file.mu

@@ -31,7 +31,7 @@ $main-body: {
         var c/eax: byte <- read-byte-buffered in-addr
         compare c, 0xffffffff  # EOF marker
         break-if-=
-        print-byte 0, c
+        print-grapheme 0, c
         loop
       }
     }