transition various bits of code to unicode.h interfaces

[disorder] / scripts / make-unidata

diff --git a/scripts/make-unidata b/scripts/make-unidata
index 009ae19273e19f3ae2cb822a613212b657ba6b03..4339fee07c527430a0bca619e74edab042b2e406 100755 (executable)
--- a/scripts/make-unidata
+++ b/scripts/make-unidata
@@ -59,7 +59,13 @@ sub key {
 # Size of a subtable
 #
 # This can be varied to trade off the number of subtables against their size.
-our $modulus = 128;
+# 16 gave the smallest results last time I checked (on a Mac with a 32-bit
+# build).
+our $modulus = 16;
+
+if(@ARGV) {
+    $modulus = shift;
+}
 
 # Where to break the table.  There is a huge empty section of the Unicode
 # code space and we deal with this by simply leaving it out of the table.
@@ -365,6 +371,11 @@ out("struct unidata {\n",
 
 out("extern const struct unidata *const unidata[];\n");
 
+out("extern const struct unicode_utf8_row {\n",
+    "  uint8_t count;\n",
+    "  uint8_t min2, max2;\n",
+    "} unicode_utf8_valid[];\n");
+
 out("#define UNICODE_NCHARS ", ($max + 1), "\n");
 out("#define UNICODE_MODULUS $modulus\n");
 out("#define UNICODE_BREAK_START $break_start\n");
@@ -521,8 +532,45 @@ for(my $base = 0; $base <= $max; $base += $modulus) {
 }
 out("};\n");
 
+out("const struct unicode_utf8_row unicode_utf8_valid[] = {\n");
+for(my $c = 0; $c <= 0x7F; ++$c) {
+    out(" { 1, 0, 0 }, /* $c */\n");
+}
+for(my $c = 0x80; $c < 0xC2; ++$c) {
+    out(" { 0, 0, 0 }, /* $c */\n");
+}
+for(my $c = 0xC2; $c <= 0xDF; ++$c) {
+    out(" { 2, 0x80, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xE0; $c <= 0xE0; ++$c) {
+    out(" { 3, 0xA0, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xE1; $c <= 0xEC; ++$c) {
+    out(" { 3, 0x80, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xED; $c <= 0xED; ++$c) {
+    out(" { 3, 0x80, 0x9F }, /* $c */\n");
+}
+for(my $c = 0xEE; $c <= 0xEF; ++$c) {
+    out(" { 3, 0x80, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xF0; $c <= 0xF0; ++$c) {
+    out(" { 4, 0x90, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xF1; $c <= 0xF3; ++$c) {
+    out(" { 4, 0x80, 0xBF }, /* $c */\n");
+}
+for(my $c = 0xF4; $c <= 0xF4; ++$c) {
+    out(" { 4, 0x80, 0x8F }, /* $c */\n");
+}
+for(my $c = 0xF5; $c <= 0xFF; ++$c) {
+    out(" { 0, 0, 0 }, /* $c */\n");
+}
+out("};\n");
+
 close STDOUT or die "unidata.c: $!\n";
 
+printf STDERR "modulus=%d\n", $modulus;
 printf STDERR "max=%04X\n", $max;
 print STDERR "subtables=$subtablecounter, subtablessaved=$subtablessaved\n";
 print STDERR "decompsaved=$decompsaved cfsaved=$cfsaved\n";