v2.0.0 - Overhauled the 'read' function in the terminal library, added a new Unicode library, and various functions added to the filesystem library.

In older versions, the 'read' function in the terminal library (termlib.lua) was most likely vibecoded by Fluxdrive, and was basically very inefficient. The new Unicode library includes functions for getting a code point from a character, and iterating Unicode characters from a string or an iterator function that returns every byte. It is now possible, in the filesystem library, to make virtual 'read streams' (filesystem.makeReadStream), which does the same thing as opening a file with some specific content. There are some new functions in read streams, which allows you to loop through bytes (open(...):iterateBytes), and loop through Unicode characters (open(...):iterateUnicodeChars). The edit app will be updated to v1.2.1 for importing the new Unicode library.

halyde/lib/unicode.lua

@@ -0,0 +1,100 @@
+local unicodeLib
+local LLunicode
+if table.copy then
+  unicodeLib = table.copy(unicode)
+  LLunicode = table.copy(unicode)
+else
+  unicodeLib = {}
+  LLunicode = unicode
+end
+
+function unicodeLib.readCodePoint(readByte)
+  local function inRange(min,max,...)
+    for _,v in ipairs({...}) do
+      if not (v and v>=min and v<max) then return false end
+    end
+    return true
+  end
+
+  local byte = readByte()
+  if byte==nil then return end
+
+  if byte < 0x80 then
+    -- ASCII character (0xxxxxxx)
+    return byte
+  elseif byte < 0xC0 then
+    -- Continuation byte (10xxxxxx), invalid at start position
+    return nil
+  elseif byte < 0xE0 then
+    -- 2-byte sequence (110xxxxx 10xxxxxx)
+    local byte2 = readByte()
+    if byte2==nil then return nil end
+    if inRange(0x80,0xC0,byte2) then
+      local code_point = ((byte & 0x1F) << 6) | (byte2 & 0x3F)
+      return code_point
+    end
+  elseif byte < 0xF0 then
+    -- 3-byte sequence (1110xxxx 10xxxxxx 10xxxxxx)
+    local byte2, byte3 = readByte(), readByte()
+    if byte2==nil and byte3==nil then return nil end
+    if inRange(0x80,0xC0,byte2,byte3)then
+      local code_point = ((byte & 0x0F) << 12) | ((byte2 & 0x3F) << 6) | (byte3 & 0x3F)
+      return code_point
+    end
+  elseif byte < 0xF8 then
+    -- 4-byte sequence (11110xxx 10xxxxxx 10xxxxxx 10xxxxxx)
+    local byte2, byte3, byte4 = readByte(), readByte(), readByte()
+    if byte2==nil and byte3==nil and byte4==nil then return nil end
+    if inRange(0x80,0xC0,byte2,byte3,byte4) then
+      local code_point = ((byte & 0x07) << 18) | ((byte2 & 0x3F) << 12) | ((byte3 & 0x3F) << 6) | (byte4 & 0x3F)
+      return code_point
+    end
+  end
+
+  -- Invalid UTF-8 byte sequence
+  return nil
+end
+
+function unicodeLib.readChar(readByte)
+  checkArg(1,readByte,"function")
+  return LLunicode.char(unicodeLib.readCodePoint(readByte))
+end
+
+function unicodeLib.codepoint(chr)
+  checkArg(1,readByte,"string")
+  local ptr = 1
+  return readUniChar(function()
+      local byte = chr:byte(ptr)
+      ptr=ptr+1
+      return byte
+  end),ptr-1
+end
+
+function unicodeLib.iterate(readByte)
+  checkArg(1,readByte,"string","function")
+  if type(readByte)=="string" then
+    local str,ptr = readByte,0
+    readByte = function()
+      ptr=ptr+1
+      return str:byte(ptr)
+    end
+  end
+  return function()
+    local point = unicodeLib.readCodePoint(readByte)
+    if point==nil then return nil end
+    return LLunicode.char(point),point
+  end
+end
+
+unicodeLib.char = LLunicode.char
+unicodeLib.charWidth = LLunicode.charWidth
+unicodeLib.isWide = LLunicode.isWide
+unicodeLib.len = LLunicode.len
+unicodeLib.lower = LLunicode.lower
+unicodeLib.reverse = LLunicode.reverse
+unicodeLib.sub = LLunicode.sub
+unicodeLib.upper = LLunicode.upper
+unicodeLib.wlen = LLunicode.wlen
+unicodeLib.wtrunc = LLunicode.wtrunc
+
+return unicodeLib