[Static if] replace overload constraints with static if (searching.d)

[wilzbach]

Split up of #5145

[wilzbach]

(some comments about the constraints)

[wilzbach]

The next overload is isNarrowString!R1 && isNarrowString!R2, which implies R2 is either a narrow string or an input range with the same element type as R1.

[wilzbach]

The next overload is isBidirectionalRange!R && is(typeof(binaryFun!pred(doesThisEnd.back, withThis)) : bool)

Which implies it's either bidirectional and back of both ends is comparable with pred or

Todo!!!

[wilzbach]

This groups these overloads:

isRandomAccessRange!R1 && hasLength!R1 && hasSlicing!R1 && isBidirectionalRange!R2 && is(typeof(binaryFun!pred(haystack.front, needle.front)) : bool))

• isRandomAccessRange!R1 && isForwardRange!R2 && !isBidirectionalRange!R2 && is(typeof(binaryFun!pred(haystack.front, needle.front)) : bool))

Notice

• for all of them the front of both ranges needs to be comparable with the given binary function: is(typeof(binaryFun!pred(haystack.front, needle.front)) : bool)) is required
• overload 2 & 3 can be grouped further as they both require RandomAccess

[wilzbach]

The next overload is isBidirectionalRange!R && is(typeof(binaryFun!pred(doesThisEnd.back, withThis)) : bool))
-> isBidirectionalRange!R is the overlapping constraint.

[UplinkCoder]

The reason to have the overloads is to have the acceptable overloads visible.

[wilzbach]

The reason to have the overloads is to have the acceptable overloads visible.

And the reason for unifying them under one function is that @WalterBright and others agree that
the user shouldn't be bothered with implementation details.

[wilzbach]

Mostly superseded by #5575