Count characters instead of bytes

[mykhailynab]

For queries such as SELECT "なにか" FROM Y WHERE "なにか" = 'test; which are using UTF-8 characters as column names, counting lengths as byte counts (which len() does) may cause bugs.

For instance, the above query would produce an error Unterminated string literal at Line: 1, Column 47, while the correct column number is 35.

So, simply use .chars().count() instead. Replacing len() with this produces correct output.

[alamb]

Thank you for the contribution @michael-2956 !

Can you please add a test to this PR (so that this feature doesn't potentially get broken by a future change without us knowing)

[coveralls]

Pull Request Test Coverage Report for Build 2629916400

• 12 of 12 (100.0%) changed or added relevant lines in 1 file are covered.
• 412 unchanged lines in 5 files lost coverage.
• Overall coverage increased (+0.1%) to 89.89%

---

Files with Coverage Reduction	New Missed Lines	%
tests/sqlparser_mysql.rs	1	99.8%
src/ast/query.rs	17	86.61%
tests/sqlparser_common.rs	27	97.15%
src/tokenizer.rs	71	89.27%
src/parser.rs	296	83.39%

Totals
Change from base Build 2535757386:	0.1%
Covered Lines:	8873
Relevant Lines:	9871

---

💛 - Coveralls

[alamb]

LGTM -- thank you @michael-2956