add @string.Index and refactor json parser

[Yu-zh]

The first commit adds the type @string.Index. Now there are three accessing methods for String:

• nth_charcode(String, Int) -> Int // access n-th charcode in O(1) time
• nth_codepoint(String, Int) -> Char // access n-th codepoint in O(n) time
• codepoint_at(String, @string.Index) -> Char // access codepoint at given index in O(1) time
  We might need another one for completeness but I'm not sure about whether it's necessary:
• charcode_at(String, @string.Index) -> Int // access charcode at given index in O(1) time
  The problem with the fourth method is that it needs another parameter shift~ : Int so that it can access the trailing surrogate.

And for taking view, there are two methods:

• op_as_view(String, Int, Int) -> @string.View // taking the view by skipping a number of codepoints on both ends
• as_view(String, @string.Index, @string.Index) -> @string.View // taking the view by just using the given index

---

The second commit shows how to use these methods to rewrite the json parser. The main benefit here is that there's no manual checking of surrogate pairs without sacrificing the efficiency.

[coveralls]

Pull Request Test Coverage Report for Build 5274

Details

• 49 of 60 (81.67%) changed or added relevant lines in 7 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage decreased (-0.06%) to 87.209%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
json/lex_number.mbt	1	2	50.0%
string/index.mbt	38	48	79.17%

Totals
Change from base Build 5273:	-0.06%
Covered Lines:	5577
Relevant Lines:	6395

---

💛 - Coveralls