Reduce unnecessary COW on Vector by make writing explicit

This commit makes operator[] on Vector const and adds a write proxy to it.  From
now on writes to Vectors need to happen through the .write proxy. So for
instance:

Vector<int> vec;
vec.push_back(10);
std::cout << vec[0] << std::endl;
vec.write[0] = 20;

Failing to use the .write proxy will cause a compilation error.

In addition COWable datatypes can now embed a CowData pointer to their data.
This means that String, CharString, and VMap no longer use or derive from
Vector.

_ALWAYS_INLINE_ and _FORCE_INLINE_ are now equivalent for debug and non-debug
builds. This is a lot faster for Vector in the editor and while running tests.
The reason why this difference used to exist is because force-inlined methods
used to give a bad debugging experience. After extensive testing with modern
compilers this is no longer the case.

1 files changed, 3 insertions, 3 deletions

diff --git a/core/ring_buffer.h b/core/ring_buffer.h
index de4757612a..00628a4ab3 100644
--- a/core/ring_buffer.h
+++ b/core/ring_buffer.h
@@ -137,7 +137,7 @@ public:
 
 	Error write(const T &p_v) {
 		ERR_FAIL_COND_V(space_left() < 1, FAILED);
-		data[inc(write_pos, 1)] = p_v;
+		data.write[inc(write_pos, 1)] = p_v;
 		return OK;
 	};
 
@@ -156,7 +156,7 @@ public:
 			int total = end - pos;
 
 			for (int i = 0; i < total; i++) {
-				data[pos + i] = p_buf[src++];
+				data.write[pos + i] = p_buf[src++];
 			};
 			to_write -= total;
 			pos = 0;
@@ -196,7 +196,7 @@ public:
 		data.resize(1 << p_power);
 		if (old_size < new_size && read_pos > write_pos) {
 			for (int i = 0; i < write_pos; i++) {
-				data[(old_size + i) & mask] = data[i];
+				data.write[(old_size + i) & mask] = data[i];
 			};
 			write_pos = (old_size + write_pos) & mask;
 		} else {