Refactor to loosen list parsing

This eases the code a bit to do exactly the same, but remove some unneeded code.
If indented code is not supported (say, someone removed the tokenizer), this
also adds support for lists that are indented more than 4 spaces.
All other tokenizers already supported that.

Author: wooorm

packages/remark-parse/lib/tokenize/list.js

@@ -36,7 +36,7 @@ function list(eat, value, silent) {
   var index = 0
   var length = value.length
   var start = null
-  var size = 0
+  var size
   var queue
   var ordered
   var character
@@ -64,21 +64,13 @@ function list(eat, value, silent) {
   while (index < length) {
     character = value.charAt(index)
 
-    if (character === tab) {
-      size += tabSize - (size % tabSize)
-    } else if (character === space) {
-      size++
-    } else {
+    if (character !== tab && character !== space) {
       break
     }
 
     index++
   }
 
-  if (size >= tabSize) {
-    return
-  }
-
   character = value.charAt(index)
 
   if (character === asterisk || character === plusSign || character === dash) {
@@ -141,7 +133,6 @@ function list(eat, value, silent) {
       nextIndex = length
     }
 
-    end = index + tabSize
     size = 0
 
     while (index < length) {
@@ -158,10 +149,6 @@ function list(eat, value, silent) {
       index++
     }
 
-    if (size >= tabSize) {
-      indented = true
-    }
-
     if (item && size >= item.indent) {
       indented = true
     }