mieszko4 · December 3, 2022 22:16 · Dec 3, 2022 · Dec 3, 2022 · Dec 3, 2022 · Dec 2, 2022
diff --git a/atob.lua b/atob.lua
@@ -6,9 +6,9 @@
 local function extractChar(unit, idx)
 	-- Skip mask for first ASCII char since shifting will destroy other two parts
 	if idx == 2 then
-		local charCode = (unit) >> (8 * idx)
+	  local charCode = (unit) >> (8 * idx)
 
-		return string.char(charCode)
+	  return string.char(charCode)
 	end
 
 	-- Extract specific 8-bit

diff --git a/atob.lua b/atob.lua
@@ -4,6 +4,13 @@
 
 -- Define function to extract 8-bit ASCII char from 24-bit base64 unit
 local function extractChar(unit, idx)
+	-- Skip mask for first ASCII char since shifting will destroy other two parts
+	if idx == 2 then
+		local charCode = (unit) >> (8 * idx)
+
+		return string.char(charCode)
+	end
+
 	-- Extract specific 8-bit
 	local mask = (2^8 - 1) << (8 * idx)
 

diff --git a/atob.lua b/atob.lua
@@ -6,19 +6,14 @@
 local function extractChar(unit, idx)
 	-- Extract specific 8-bit
 	local mask = (2^8 - 1) << (8 * idx)
-
+  
 	-- Move 8-bit to the beginning so it becomes a char code
 	local charCode = (unit & mask) >> (8 * idx)
 
-	-- Return empty string in case of padding
-	if charCode == 0 then
-		return ""
-	end
-
 	-- Convert to string
 	return string.char(charCode)
   end
-
+  
   -- Define mapping of base64 character to number
   local mapping = {
 	  A=0, B=1, C=2, D=3, E=4, F=5, G=6, H=7,
@@ -30,9 +25,11 @@ local function extractChar(unit, idx)
 	  w=48, x=49, y=50, z=51, ["0"]=52, ["1"]=53, ["2"]=54, ["3"]=55,
 	  ["4"]=56, ["5"]=57, ["6"]=58, ["7"]=59, ["8"]=60, ["9"]=61, ["+"]=62, ["/"]=63
   }
-
+  local padding = "="
+  local missing = ""
+
   local function atob(encodedString)
-    -- Get four 6-bit chunks
+    -- Get four 6-bit chunks which forms one unit
   	-- There must be at least two 6-bit chunks to cover at least one 8-bit chunk
 	local decodedString = encodedString:gsub("(.)(.)(.?)(.?)", function(c1, c2, c3, c4)
 	  -- Convert each 6-bit chunk to number
@@ -44,14 +41,18 @@ local function extractChar(unit, idx)
 	  -- Set optional chunks to 0 to be able to multiplex
 	  if b3 == nil then b3 = 0 end
 	  if b4 == nil then b4 = 0 end
-
+  
 	  -- Multiplex onto 24-bit number
 	  local muxedUnit = (b1 << (6 * 3)) + (b2 << (6 * 2)) + (b3 << (6 * 1)) + (b4 << (6 * 0))
-
+	  
 	  -- Extract 8-bit chunks
+	  -- Use empty string if unit is not full or it is padded
+	  -- and handle A base64 characters at the end of encodedString
 	  local r1 = extractChar(muxedUnit, 2)
-	  local r2 = extractChar(muxedUnit, 1)
-	  local r3 = extractChar(muxedUnit, 0)
+	  -- if c3 is padding or missing then we can skip r2
+	  local r2 = (c3 == padding or c3 == missing) and "" or extractChar(muxedUnit, 1)
+	  -- if c4 is padding or missing then we can skip r3
+	  local r3 = (c4 == padding or c4 == missing) and "" or extractChar(muxedUnit, 0)
 
 	  -- Concatanate
 	  return r1..r2..r3
@@ -62,7 +63,9 @@ end
 
 return atob
 
+-- Test plan
 assert(atob("bGlnaHQgd28=") == "light wo", "full units, single padding")
+assert(atob("bGlnaHQgd28A") == "light wo\x00", "full units, explicit A")
 assert(atob("bGlnaHQgd28") == "light wo", "2/3 of unit")
 assert(atob("bGlnaHQgdw==") == "light w", "full unit, double padding")
 assert(atob("bGlnaHQgdw") == "light w", "1/3 of unit")

diff --git a/atob.lua b/atob.lua
@@ -0,0 +1,70 @@
+-- Lua uses at least 8 bytes to represent a number
+-- Base64 minimum processable unit is represented by four 6-bit chunks which is 24-bits => 3 bytes
+-- Hence we can safely use bit operations on number type for multiplexing and extraction
+
+-- Define function to extract 8-bit ASCII char from 24-bit base64 unit
+local function extractChar(unit, idx)
+	-- Extract specific 8-bit
+	local mask = (2^8 - 1) << (8 * idx)
+
+	-- Move 8-bit to the beginning so it becomes a char code
+	local charCode = (unit & mask) >> (8 * idx)
+
+	-- Return empty string in case of padding
+	if charCode == 0 then
+		return ""
+	end
+
+	-- Convert to string
+	return string.char(charCode)
+  end
+
+  -- Define mapping of base64 character to number
+  local mapping = {
+	  A=0, B=1, C=2, D=3, E=4, F=5, G=6, H=7,
+	  I=8, J=9, K=10, L=11, M=12, N=13, O=14, P=15,
+	  Q=16, R=17, S=18, T=19, U=20, V=21, W=22, X=23,
+	  Y=24, Z=25, a=26, b=27, c=28, d=29, e=30, f=31,
+	  g=32, h=33, i=34, j=35, k=36, l=37, m=38, n=39,
+	  o=40, p=41, q=42, r=43, s=44, t=45, u=46, v=47,
+	  w=48, x=49, y=50, z=51, ["0"]=52, ["1"]=53, ["2"]=54, ["3"]=55,
+	  ["4"]=56, ["5"]=57, ["6"]=58, ["7"]=59, ["8"]=60, ["9"]=61, ["+"]=62, ["/"]=63
+  }
+
+  local function atob(encodedString)
+    -- Get four 6-bit chunks
+  	-- There must be at least two 6-bit chunks to cover at least one 8-bit chunk
+	local decodedString = encodedString:gsub("(.)(.)(.?)(.?)", function(c1, c2, c3, c4)
+	  -- Convert each 6-bit chunk to number
+	  local b1 = mapping[c1] --24,23,22,21,20,19
+	  local b2 = mapping[c2] --18,17,16,15,14,13
+	  local b3 = mapping[c3] --12,11,10,09,08,07
+	  local b4 = mapping[c4] --06,05,04,03,02,01
+
+	  -- Set optional chunks to 0 to be able to multiplex
+	  if b3 == nil then b3 = 0 end
+	  if b4 == nil then b4 = 0 end
+
+	  -- Multiplex onto 24-bit number
+	  local muxedUnit = (b1 << (6 * 3)) + (b2 << (6 * 2)) + (b3 << (6 * 1)) + (b4 << (6 * 0))
+
+	  -- Extract 8-bit chunks
+	  local r1 = extractChar(muxedUnit, 2)
+	  local r2 = extractChar(muxedUnit, 1)
+	  local r3 = extractChar(muxedUnit, 0)
+
+	  -- Concatanate
+	  return r1..r2..r3
+  end)
+
+  return decodedString
+end
+
+return atob
+
+assert(atob("bGlnaHQgd28=") == "light wo", "full units, single padding")
+assert(atob("bGlnaHQgd28") == "light wo", "2/3 of unit")
+assert(atob("bGlnaHQgdw==") == "light w", "full unit, double padding")
+assert(atob("bGlnaHQgdw") == "light w", "1/3 of unit")
+assert(atob("bGlnaHQgd28=") == "light wo", "full units, single padding")
+assert(atob("bGlnaHQgd29y") == "light wor", "full units, no padding")