Improve unicode finder script

bin/cp2utf8

@@ -0,0 +1,34 @@
+#!/usr/bin/env luajit
+
+local bit = require 'bit'
+
+local function code_to_utf8(code)
+  if code <= 0x7f then
+    return string.char(code)
+  elseif code <= 0x7ff then
+    return string.char(
+      bit.bor(0xC0, bit.rshift(code, 6)),
+      bit.bor(0x80, bit.band(code, 0x3f))
+    )
+  elseif code <= 0xffff then
+    return string.char(
+      bit.bor(0xE0, bit.rshift(code, 12)),
+      bit.bor(0x80, bit.band(bit.rshift(code, 6), 0x3f)),
+      bit.bor(0x80, bit.band(code, 0x3f))
+    )
+  elseif code <= 0x10FFFF then
+    return string.char(
+      bit.bor(0xE0, bit.rshift(code, 18)),
+      bit.bor(0x80, bit.band(bit.rshift(code, 12), 0x3f)),
+      bit.bor(0x80, bit.band(bit.rshift(code, 06), 0x3f)),
+      bit.bor(0x80, bit.band(code, 0x3f))
+    )
+  else
+    error 'Invalid character!'
+  end
+end
+
+for line in io.stdin:lines() do
+  local code = tonumber(line:match('%x+'), 16)
+  print(line, code_to_utf8(code))
+end

bin/unicode

@@ -5,10 +5,11 @@ then
   echo $namelist not found, downloading...
   mkdir -p $(dirname "$namelist")
   touch "$namelist"
-  wget 'https://www.unicode.org/Public/UCD/latest/ucd/NamesList.txt' \
-  -O "$namelist" || exit
+  curl 'https://www.unicode.org/Public/UCD/latest/ucd/NamesList.txt' \
+  | sed -n -e '/<.*>/d' -e '/^[0-9A-F]\+/p' \
+  | cp2utf8 > "$namelist" || exit
 fi
-codepoint=$(grep -i '^[0-9A-F]\+' "$namelist" | sed -e '/<.*>/d' -e 's/\t/ /g' | dmenu -i -l 20 -b | awk '{print $1}' | grep '^[0-9A-F]\+')
+codepoint=$(cat "$namelist" | sed -e 's/\t/   /g' | dmenu -i -l 20 -b | awk '{print $1}' | grep '^[0-9A-F]\+')
 if [ -z "$codepoint" ];
 then echo No codepoint selected >&2; exit
 fi