update readme

README.md

@@ -15,4 +15,54 @@ But also optimized the constant input for GCC/Clang that supports __builtin_cons
 TODO : \
 AVX-512 optimaztion for addc_n/subc_n.  \
 clz, ctz, popcount, mul_1, mul_n, ... \
-constexpr math functions ...
+constexpr math functions ...
+
+3. preprocessor \
+WJR_PP_ADD, WJR_PP_SUB, WJR_PP_(LT|LE|GT|GE|EQ|NE), WJR_PP_INC, WJR_PP_DEC, WJR_PP_NEG. \
+Currently, only 6-bit operations are supported  \
+WJR_PP_BOOL_IF, WJR_PP_BOOL_NOT, WJR_PP_BOOL_AND, WJR_PP_BOOL_OR, WJR_PP_BOOL_XOR.  \
+WJR_PP_QUEUE :  \
+A queue is similar to (a, b, c, d, e).    \
+WJR_PP_QUEUE_CALL_N(queue, ops, n) : Each time the first two elements of the queue are evaluated   \
+according to the first element of ops and placed back at the head of the queue.     \
+For example : 
+```
+#define F(a, b) a + b
+WJR_PP_QUEUE_CALL_N((1, 2, 3), (F, F), 2) 
+
+init :
+    queue : (1, 2, 3), ops : (F, F)
+first time :
+    queue : (1 + 2, 3), ops : (F)
+second time :
+    queue : (1 + 2 + 3), ops : ()
+```
+In this way, it can be easily extended to various functions.    \
+But it is difficult to implement a recursive version.   \
+I will try to support recursion in the future.
+support now : 
+WJR_PP_QUEUE_TRANSFORM(queue, op) : (a, b, c), f -> (f(a), f(b), f(c))
+WJR_PP_QUEUE_TRANSFORMS(queue, ops) : (a, b, c), (f, g, h) -> (f(a), g(b), h(c))
+WJR_PP_QUEUE_ACCUMULATE(queue, op) : init, (a, b, c), f -> f(f(f(0, 1), 2), 3)
+WJR_PP_QUEUE_ACCUMULATES(queue, op) : init, (a, b, c), (f, g, h) -> h(g(f(0, 1), 2), 3)
+...
+
+4. compressed_pair
+Currently undergoing refactoring. I hope it can be designed as compressed_tuple.
+
+5. vector
+vector is broken now.   \
+Vector is designed to have customizable internal data structures,   \
+requiring only partial implementation of necessary interfaces.  \
+The biggest drawback of a vector is that in some cases it knows the actual size,    \
+and in some performance critical positions, the vector is not very suitable. \
+The vector designed based on STL is expected to easily expand to this \
+fixed length capacity situation, throwing an exception when the expansion \
+exceeds the fixed length capacity instead of reallocating. \
+And it should be able to be extended to SSO vectors. \
+And I hope to expand to more complex situations in the future, \
+such as copy on write.
+
+6. more template operations
+
+7. more constexpr support