Rename isupper, islower, ucfirst and lcfirst

[nalimilan]

Spell "uppercase" and "lowercase" in full for consistency with uppercase and lowercase functions.

The current function names seem to be inspired from different languages (Python and Perl), and are therefore inconsistent. Here's a small survey of what other languages use:

This PR:	uppercase	lowercase	isuppercase	islowercase	uppercasefirst	lowercasefirst
Python	upper	lower	isupper	islower	capitalize
Ruby	upcase	downcase			capitalize
Perl	uc	lc			ucfirst	lcfirst
Java	toUpperCase	toLowerCase	isUpperCase	isLowerCase	StringUtils.capitalize	StringUtils.uncapitalize
Rust	to_uppercase	to_lowercase	is_uppercase	is_lowercase
Swift	uppercased	lowercased	long*	long*
R	toupper	tolower			Hmisc::capitalize

* long: CharacterSet.uppercaseLetters.contains(c) and CharacterSet.lowercaseLetters.contains(c).

[JeffBezanson]

Since we use C names for character predicates, maybe we should keep isupper and use toupper and tolower.

[JeffBezanson]

The history file should not be changed.

[nalimilan]

Since we use C names for character predicates, maybe we should keep isupper and use toupper and tolower.

That's a possibility and it would give shorter identifiers. However the "to" prefix doesn't convey very useful information, and AFAIK we don't use it anywhere else.

[nalimilan]

Another advantage of using "case" everywhere is that it's consistent with the titlecase function, which can hardly be renamed to title.

[StefanKarpinski]

We're increasingly deviating from the C names for these and the C names are pretty bad to begin with. I think we should just pick a good set of names for these and be guided by other high-level languages like Python and Ruby (and Swift, etc.).

[JeffBezanson]

Ok, I'm fine with the names in this PR. Just throwing out some options.

[coveralls]

Coverage decreased (-60.5%) to 30.381% when pulling f79b469 on nl/case into f354532 on master.