Previously, "Foo" would match "Foo" but not "FOO". Now it matches both. Bottom
line, an uppercase char matches only uppercase chars but lowercase chars can
match either. We used to insist that lowercase chars in the query match
lowercase chars in the candidate strings IF the query also had uppercase
letters; this was dumb and a poor user experience.
Fixes#501.
GetWordBoundaryChars would return "allcaps" for "ALL_CAPS" instead of "ac". This
would manifest as ranking (for instance) "STDIN_FILENO" ahead of "stdin" for
query "std", which is terrible.
This bug has been present in YCM for many months, but no one noticed the issue.
Fixes#272.
The point is that we want to prefer candidates that have the query characters
"earlier" in their text, e.g. "xxabcxxx" over "xxxxxabc" for "abc" query.
