Improved str8__fmt_kt1l (watl.v0.msvc.c), need to propogate change to other versions

Prepparing to explore Lottes's C-- more, also how crufty it looks in odin.

C/watl.v0.llvm.lottes.c

@@ -0,0 +1,61 @@
+/*
+WATL Exercise
+Version:   0 (From Scratch, 1-Stage Compilation, LLVM & WinAPI Only, Win CRT Multi-threaded Static Linkage)
+Host:      Windows 11 (x86-64)
+Toolchain: LLVM (2025-08-30), C-Stanard: 11
+
+Following strictly: Neokineogfx - Fixing C
+https://youtu.be/RrL7121MOeA
+*/
+
+#pragma clang diagnostic ignored "-Wunused-const-variable"
+#pragma clang diagnostic ignored "-Wunused-but-set-variable"
+#pragma clang diagnostic ignored "-Wswitch"
+#pragma clang diagnostic ignored "-Wunused-variable"
+#pragma clang diagnostic ignored "-Wunknown-pragmas"
+#pragma clang diagnostic ignored "-Wvarargs"
+#pragma clang diagnostic ignored "-Wunused-function"
+#pragma clang diagnostic ignored "-Wbraced-scalar-init"
+#pragma clang diagnostic ignored "-W#pragma-messages"
+#pragma clang diagnostic ignored "-Wstatic-in-inline"
+#pragma clang diagnostic ignored "-Wkeyword-macro"
+#pragma clang diagnostic ignored "-Wc23-compat"
+#pragma clang diagnostic ignored "-Wreserved-identifier"
+#pragma clang diagnostic ignored "-Wpre-c11-compat"
+#pragma clang diagnostic ignored "-Wc23-extensions"
+#pragma clang diagnostic ignored "-Wunused-macros"
+#pragma clang diagnostic ignored "-Wdeclaration-after-statement"
+#pragma clang diagnostic ignored "-Wunsafe-buffer-usage"
+#pragma clang diagnostic ignored "-Wc++-keyword"
+#pragma clang diagnostic ignored "-Wimplicit-function-declaration"
+#pragma clang diagnostic ignored "-Wcast-align"
+#pragma clang diagnostic ignored "-Wunused-parameter"
+#pragma clang diagnostic ignored "-Wswitch-default"
+#pragma clang diagnostic ignored "-Wmissing-field-initializers"
+#pragma clang diagnostic ignored "-Wgnu-zero-variadic-macro-arguments"
+#pragma clang diagnostic ignored "-Wpointer-sign"
+
+#pragma region Header
+
+#pragma region DSL
+#define A_(x)   __attribute__((aligned (x)))
+#define E_(x,y) __builtin_expect(x,y)
+#define S_      static
+#define I_      static inline __attribute__((always_inline))
+#define N_      static        __attribute__((noinline))
+#define R_      __restrict                                     
+#define V_      volatile                                       
+#define W_      __attribute((__stdcall__)) __attribute__((__force_align_arg_pointer__))
+#pragma endregion DSL
+
+#pragma endregion Header
+
+#pragma region Implementation
+#pragma endregion Implementation
+
+int main(void)
+{
+	return 0;
+}
+
+#pragma clang diagnostic pop

C/watl.v0.llvm.lottes_hybrid.c

@@ -4,7 +4,7 @@ Version:   0 (From Scratch, 1-Stage Compilation, LLVM & WinAPI Only, Win CRT Mul
 Host:      Windows 11 (x86-64)
 Toolchain: LLVM (2025-08-30), C-Stanard: 11
 
-Based on: Neokineogfx - Fixing C
+Based on: Neokineogfx - Fixing C, personalized to include typeinfo more readily.
 https://youtu.be/RrL7121MOeA
 */
 
@@ -39,21 +39,6 @@ https://youtu.be/RrL7121MOeA
 #pragma region Header
 
 #pragma region DSL
-#if 0
-// Original macros
-
-#define A_(x)   __attribute__((aligned (x)))
-#define E_(x,y) __builtin_expect(x,y)
-#define S_      static
-#define I_      static inline __attribute__((always_inline))
-#define N_      static        __attribute__((noinline))
-#define R_      __restrict                                     
-#define V_      volatile                                       
-#define W_      __attribute((__stdcall__)) __attribute__((__force_align_arg_pointer__))
-#endif
-
-// Ones I'm deciding to use..
-
 #define align_(value)     __attribute__((aligned (value)))             // for easy alignment
 #define expect_(x, y)     __builtin_expect(x, y)                       // so compiler knows the common path
 #define finline           static inline __attribute__((always_inline)) // force inline
@@ -130,6 +115,9 @@ enum { false = 0, true  = 1, true_overflow, };
 #define offset_of(type, member)             cast(U8, & (((type*) 0)->member))
 #define size_of(data)                       cast(U8, sizeof(data))
 
+#define r_(ptr)                             cast(typeof_ptr(ptr)*R_, ptr)
+#define v_(ptr)                             cast(typeof_ptr(ptr)*V_, ptr)
+
 #define kilo(n)                             (cast(U8, n) << 10)
 #define mega(n)                             (cast(U8, n) << 20)
 #define giga(n)                             (cast(U8, n) << 30)
@@ -158,17 +146,17 @@ def_signed_ops(ge, >=) def_signed_ops(le, <=)
 #define ge_s(a,b)  def_generic_sop(ge, a,b)
 #define le_s(a,b)  def_generic_sop(le, a,b)
 
-finline U4 AtmAdd_u4 (U4_R a, U4 v){__asm__ volatile("lock xaddl %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
-finline U8 AtmAdd_u8 (U8_R a, U8 v){__asm__ volatile("lock xaddq %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
-finline U4 AtmSwap_u4(U4_R a, U4 v){__asm__ volatile("lock xchgl %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
-finline U8 AtmSwap_u8(U8_R a, U8 v){__asm__ volatile("lock xchgq %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
+finline U4 atm_add_u4 (U4_R a, U4 v){__asm__ volatile("lock xaddl %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
+finline U8 atm_add_u8 (U8_R a, U8 v){__asm__ volatile("lock xaddq %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
+finline U4 atm_swap_u4(U4_R a, U4 v){__asm__ volatile("lock xchgl %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
+finline U8 atm_swap_u8(U8_R a, U8 v){__asm__ volatile("lock xchgq %0,%1":"=r"(v),"=m"(*a):"0"(v),"m"(*a):"memory","cc");return v;}
 #pragma endregion DSL
 
 #pragma region Strings
 typedef unsigned char def_tset(UTF8);
-typedef def_struct(Str8)       { UTF8*R_ ptr; U8 len; }; typedef Str8 def_tset(Slice_UTF8);
-typedef def_struct(Slice_Str8) { Str8*R_ ptr; U8 len; };
-#define lit(string_literal) (Str8){ (UTF8*R_) string_literal, size_of(string_literal) - 1 }
+typedef def_struct(Str8)       { UTF8* ptr; U8 len; }; typedef Str8 def_tset(Slice_UTF8);
+typedef def_struct(Slice_Str8) { Str8* ptr; U8 len; };
+#define lit(string_literal)    (Str8){ (UTF8*) string_literal, size_of(string_literal) - 1 }
 #pragma endregion Strings
 
 #pragma region Debug
@@ -217,10 +205,10 @@ U8 mem_copy            (U8 dest, U8 src, U8 length);
 U8 mem_copy_overlapping(U8 dest, U8 src, U8 length);
 B4 mem_zero            (U8 dest, U8 length);
  
-finline void BarC(void){__asm__ volatile("::""memory");} // Compiler Barrier
-finline void BarM(void){__builtin_ia32_mfence();}        // Memory   Barrier
-finline void BarR(void){__builtin_ia32_lfence();}        // Read     Barrier
-finline void BarW(void){__builtin_ia32_sfence();}        // Write    Barrier
+finline void barrier_compiler(void){__asm__ volatile("::""memory");} // Compiler Barrier
+finline void barrier_memory  (void){__builtin_ia32_mfence();}        // Memory   Barrier
+finline void barrier_read    (void){__builtin_ia32_lfence();}        // Read     Barrier
+finline void barrier_write   (void){__builtin_ia32_sfence();}        // Write    Barrier
 
 #define check_nil(nil, p) ((p) == 0 || (p) == nil)
 #define set_nil(nil, p)   ((p) = nil)
@@ -241,7 +229,7 @@ finline void BarW(void){__builtin_ia32_sfence();}        // Write    Barrier
 )
 #define sll_queue_push_n(f, l, n, next) sll_queue_push_nz(0, f, l, n, next)
 
-#define def_Slice(type)           def_struct(tmpl(Slice,type)) { type*R_ ptr; U8 len; }; typedef def_ptr_set(tmpl(Slice,type))
+#define def_Slice(type)           def_struct(tmpl(Slice,type)) { type* ptr; U8 len; }; typedef def_ptr_set(tmpl(Slice,type))
 #define slice_assert(slice)       do { assert((slice).ptr != 0); assert((slice).len > 0); } while(0)
 #define slice_end(slice)          ((slice).ptr + (slice).len)
 #define size_of_slice_type(slice) size_of( (slice).ptr[0] )
@@ -252,7 +240,7 @@ typedef def_struct(Slice_Mem) { U8 ptr; U8 len; };
 
 typedef def_Slice(void);
 typedef def_Slice(B1);
-#define slice_byte(slice) ((Slice_B1){cast(B1_R, (slice).ptr), (slice).len * size_of_slice_type(slice)})
+#define slice_byte(slice) ((Slice_B1){cast(B1*, (slice).ptr), (slice).len * size_of_slice_type(slice)})
 #define slice_fmem(mem)   slice_mem(u8_(mem), size_of(mem))
 
 finline void slice__copy(Slice_B1 dest, U8 dest_typewidth, Slice_B1 src, U8 src_typewidth);
@@ -386,8 +374,8 @@ finline Slice_Mem mem__shrink(AllocatorInfo ainfo, Slice_Mem mem, U8 size, Opts_
 #define mem_resize(ainfo, mem, size, ...) mem__resize(ainfo, mem, size, opt_args(Opts_mem_resize, __VA_ARGS__))
 #define mem_shrink(ainfo, mem, size, ...) mem__shrink(ainfo, mem, size, opt_args(Opts_mem_shrink, __VA_ARGS__))
 
-#define alloc_type(ainfo, type, ...)       (type*R_)                    mem__alloc(ainfo, size_of(type),        opt_args(Opts_mem_alloc, __VA_ARGS__)).ptr
-#define alloc_slice(ainfo, type, num, ...) (tmpl(Slice,type)){ (type*R_)mem__alloc(ainfo, size_of(type) * num,  opt_args(Opts_mem_alloc, __VA_ARGS__)).ptr, num }
+#define alloc_type(ainfo, type, ...)       (type*)                    mem__alloc(ainfo, size_of(type),        opt_args(Opts_mem_alloc, __VA_ARGS__)).ptr
+#define alloc_slice(ainfo, type, num, ...) (tmpl(Slice,type)){ (type*)mem__alloc(ainfo, size_of(type) * num,  opt_args(Opts_mem_alloc, __VA_ARGS__)).ptr, num }
 #pragma endregion Allocator Interface
 
 #pragma region FArena (Fixed-Sized Arena)
@@ -420,8 +408,8 @@ cast(type*R_, farena__push(arena, size_of(type), 1, opt_args(Opts_farena, lit(st
 #pragma endregion FArena
 
 #pragma region OS
-finline U8   Clk  (void){U8 aa,dd;__asm__ volatile("rdtsc":"=a"(aa),"=d"(dd));return aa;}
-finline void Pause(void){__asm__ volatile("pause":::"memory");}
+finline U8   clock(void){U8 aa,dd;__asm__ volatile("rdtsc":"=a"(aa),"=d"(dd));return aa;}
+finline void pause(void){__asm__ volatile("pause":::"memory");}
 
 typedef def_struct(OS_SystemInfo) {
 	U8 target_page_size;
@@ -463,7 +451,7 @@ typedef def_struct(Opts_varena_make) {
 	VArenaFlags flags;
 	A4_B1 _PAD_;
 };
-VArena_R varena__make(Opts_varena_make*R_ opts);
+VArena*   varena__make(Opts_varena_make*R_ opts);
 #define   varena_make(...) varena__make(opt_args(Opts_varena_make, __VA_ARGS__))
 
         Slice_Mem   varena__push  (VArena_R arena, U8 amount, U8 type_width, Opts_varena*R_ opts);
@@ -479,7 +467,7 @@ void varena_allocator_proc(AllocatorProc_In in, AllocatorProc_Out_R out);
 #define varena_push_mem(arena, amount, ...) varena__push(arena, amount, 1, opt_args(Opts_varena, lit(stringify(B1)), __VA_ARGS__))
 
 #define varena_push(arena, type, ...) \
-cast(type*R_, varena__push(arena, 1, size_of(type), opt_args(Opts_varena, lit(stringify(type)), __VA_ARGS__) ).ptr)
+cast(type*, varena__push(arena, 1, size_of(type), opt_args(Opts_varena, lit(stringify(type)), __VA_ARGS__) ).ptr)
 
 #define varena_push_array(arena, type, amount, ...) \
 (tmpl(Slice,type)){ varena__push(arena, size_of(type), amount, opt_args(Opts_varena, lit(stringify(type)), __VA_ARGS__)).ptr, amount }
@@ -492,16 +480,16 @@ typedef def_enum(U4, ArenaFlags) {
 	ArenaFlag_NoChain      = (1 << 1),
 };
 typedef def_struct(Arena) {
-	VArena_R   backing;
-	Arena_R    prev;
-	Arena_R    current;
+	VArena*    backing;
+	Arena*     prev;
+	Arena*     current;
 	U8         base_pos;
 	U8         pos;
 	ArenaFlags flags;
 	A4_B1 _PAD_;
 };
 typedef Opts_varena_make Opts_arena_make;
-        Arena_R     arena__make  (Opts_arena_make*R_ opts);
+        Arena*      arena__make  (Opts_arena_make*R_ opts);
         Slice_Mem   arena__push  (Arena_R arena, U8 amount, U8 type_width, Opts_arena*R_ opts);
 finline void        arena_release(Arena_R arena);
 finline void        arena_reset  (Arena_R arena);
@@ -580,7 +568,7 @@ def_struct(tmpl(KT1CX_Slot,type)) { \
 #define def_KT1CX_Cell(type, depth)    \
 def_struct(tmpl(KT1CX_Cell,type)) {    \
 	tmpl(KT1CX_Slot,type)  slots[depth]; \
-	tmpl(KT1CX_Slot,type)*R_ next;         \
+	tmpl(KT1CX_Slot,type)* next;         \
 }
 #define def_KT1CX(type)                  \
 def_struct(tmpl(KT1CX,type)) {           \
@@ -651,7 +639,7 @@ typedef def_Slice(A2_Str8);
 typedef def_KT1L_Slot(Str8);
 typedef def_KT1L(Str8);
 
-finline Str8 str8__fmt_backed(AllocatorInfo tbl_backing, AllocatorInfo buf_backing, Str8 fmt_template, Slice_A2_Str8* entries);
+finline Str8 str8__fmt_backed(AllocatorInfo tbl_backing, AllocatorInfo buf_backing, Str8 fmt_template, Slice_A2_Str8*R_ entries);
 #define str8_fmt_backed(tbl_backing, buf_backing, fmt_template, ...) \
 str8__fmt_backed(tbl_backing, buf_backing, lit(fmt_template), slice_arg_from_array(A2_Str8, __VA_ARGS__))
 
@@ -692,7 +680,7 @@ finline Str8 cache_str8(Str8Cache_R cache, Str8 str);
 
 typedef def_struct(Str8Gen) {
 	AllocatorInfo backing;
-	UTF8_R ptr;
+	UTF8* ptr;
 	U8    len;
 	U8    cap;
 };
@@ -722,7 +710,7 @@ void api_file_read_contents(FileOpInfo_R result, Str8 path, Opts_read_file_conte
 void file_write_str8       (Str8 path, Str8 content);
 
 finline FileOpInfo file__read_contents(Str8 path, Opts_read_file_contents*R_ opts);
-#define file_read_contents(path, ...) file__read_contents(path, &(Opts_read_file_contents){__VA_ARGS__})
+#define file_read_contents(path, ...) file__read_contents(path, opt_args(Opts_read_file_contents, __VA_ARGS__))
 #pragma endregion File System
 
 #pragma region WATL
@@ -743,13 +731,13 @@ typedef def_struct(WATL_Pos) {
 	S4 column;
 };
 typedef def_struct(WATL_LexMsg) {
-	WATL_LexMsg_R next;
+	WATL_LexMsg* next;
 	Str8         content;
-	WATL_Tok_R    tok;
+	WATL_Tok*    tok;
 	WATL_Pos     pos;
 };
 typedef def_struct(WATL_LexInfo) {
-	WATL_LexMsg_R  msgs;
+	WATL_LexMsg*   msgs;
 	Slice_WATL_Tok toks;
 	WATL_LexStatus signal;
 	A4_B1 _PAD_;
@@ -764,17 +752,17 @@ typedef def_struct(Opts_watl_lex) {
 };
 void         api_watl_lex(WATL_LexInfo_R info, Str8 source, Opts_watl_lex*R_ opts);
 WATL_LexInfo watl__lex   (                     Str8 source, Opts_watl_lex*R_ opts);
-#define watl_lex(source, ...) watl__lex(source, &(Opts_watl_lex){__VA_ARGS__})
+#define watl_lex(source, ...) watl__lex(source, opt_args(Opts_watl_lex, __VA_ARGS__))
 
 typedef Str8 WATL_Node; typedef def_ptr_set(WATL_Node);
 typedef def_Slice(WATL_Node);
 typedef Slice_WATL_Node def_tset(WATL_Line);
 typedef def_Slice(WATL_Line);
 typedef def_struct(WATL_ParseMsg) {
-	WATL_ParseMsg_R next;
+	WATL_ParseMsg* next;
 	Str8        content;
-	WATL_Line_R line;
-	WATL_Tok_R  tok;
+	WATL_Line*  line;
+	WATL_Tok*   tok;
 	WATL_Pos    pos;
 };
 typedef def_enum(U4, WATL_ParseStatus) {
@@ -782,7 +770,7 @@ typedef def_enum(U4, WATL_ParseStatus) {
 };
 typedef def_struct(WATL_ParseInfo) {
 	Slice_WATL_Line  lines;
-	WATL_ParseMsg_R  msgs;
+	WATL_ParseMsg*   msgs;
 	WATL_ParseStatus signal;
 	A4_B1 _PAD_;
 };
@@ -790,13 +778,13 @@ typedef def_struct(Opts_watl_parse) {
 	AllocatorInfo ainfo_msgs;
 	AllocatorInfo ainfo_nodes;
 	AllocatorInfo ainfo_lines;
-	Str8Cache_R   str_cache;
+	Str8Cache*    str_cache;
 	B4 failon_slice_constraint_fail;
 	A4_B1 _PAD_;
 };
 void           api_watl_parse(WATL_ParseInfo_R info, Slice_WATL_Tok tokens, Opts_watl_parse*R_ opts);
 WATL_ParseInfo watl__parse   (                       Slice_WATL_Tok tokens, Opts_watl_parse*R_ opts);
-#define watl_parse(tokens, ...) watl__parse(tokens, &(Opts_watl_parse){__VA_ARGS__})
+#define watl_parse(tokens, ...) watl__parse(tokens, opt_args(Opts_watl_parse, __VA_ARGS__))
 
 Str8 watl_dump_listing(AllocatorInfo buffer, Slice_WATL_Line lines);
 #pragma endregion WATL
@@ -1148,7 +1136,7 @@ inline void  os_vmem_release(U8 vm, U8 size) { VirtualFree(cast(MS_LPVOID, vm),
 
 #pragma region VArena (Virutal Address Space Arena)
 inline
-VArena_R varena__make(Opts_varena_make*R_ opts) {
+VArena* varena__make(Opts_varena_make*R_ opts) {
 	assert(opts != nullptr);
 	if (opts->reserve_size == 0) { opts->reserve_size = mega(64); }
 	if (opts->commit_size  == 0) { opts->commit_size  = mega(64); }
@@ -1159,8 +1147,8 @@ VArena_R varena__make(Opts_varena_make*R_ opts) {
 	assert(base != 0);
 	os_vmem_commit(base, commit_size, .no_large_pages = no_large_pages);
 	U8 header_size = align_pow2(size_of(VArena), MEMORY_ALIGNMENT_DEFAULT);
-	VArena_R vm = cast(VArena_R, base);
-	vm[0] = (VArena){
+	VArena* vm = cast(VArena*, base);
+	r_(vm)[0] = (VArena){
 		.reserve_start = base + header_size,
 		.reserve       = reserve_size,
 		.commit_size   = commit_size,
@@ -1294,13 +1282,13 @@ void varena_allocator_proc(AllocatorProc_In in, AllocatorProc_Out* out)
 
 #pragma region Arena (Chained Arena)
 inline
-Arena_R arena__make(Opts_arena_make*R_ opts) {
+Arena* arena__make(Opts_arena_make*R_ opts) {
 	assert(opts != nullptr);
 	U8 header_size  = align_pow2(size_of(Arena), MEMORY_ALIGNMENT_DEFAULT);
 	VArena_R current = varena__make(opts);
 	assert(current != nullptr);
-	Arena_R arena = varena_push(current, Arena);
-	arena[0] = (Arena){
+	Arena* arena = varena_push(current, Arena);
+	r_(arena)[0] = (Arena){
 		.backing  = current,
 		.prev     = nullptr,
 		.current  = arena,

C/watl.v0.msvc.c

@@ -16,78 +16,62 @@ Toolchain: MSVC 19.43, C-Stanard: 11
 #pragma region Header
 
 #pragma region DSL
-typedef unsigned __int8  U8;
-typedef signed   __int8  S8;
-typedef unsigned __int16 U16;
-typedef signed   __int16 S16;
-typedef unsigned __int32 U32;
-typedef signed   __int32 S32;
-typedef unsigned __int64 U64;
-typedef signed   __int64 S64;
-typedef unsigned char    Byte;
-typedef unsigned __int64 USIZE;
-typedef          __int64 SSIZE;
-typedef S8               B8;
-typedef S16              B16;
-typedef S32              B32;
-enum {
-	false = 0,
-	true  = 1,
-	true_overflow,
-};
 #define glue_impl(A, B)                     A ## B
 #define glue(A, B)                          glue_impl(A, B)
 #define stringify_impl(S)                   #S
 #define stringify(S)                        stringify_impl(S)
 #define tmpl(prefix, type)                  prefix ## _ ## type
 
+#define local_persist                       static
+#define global                              static
+#define internal                            static
+
+#define static_assert                       _Static_assert
+#define typeof                              __typeof__
+#define typeof_ptr(ptr)                     typeof(ptr[0])
+#define typeof_same(a, b)                   _Generic((a), typeof((b)): 1, default: 0)
+
+typedef unsigned __int8  U8;    typedef unsigned __int16 U16; typedef unsigned __int32 U32; typedef unsigned __int64 U64;
+typedef signed   __int8  S8;    typedef signed   __int16 S16; typedef signed   __int32 S32; typedef signed   __int64 S64;
+typedef unsigned char    Byte;  typedef S8               B8;  typedef S16              B16; typedef S32              B32;
+typedef unsigned __int64 USIZE; typedef __int64 SSIZE;
+typedef float            F32;   typedef double  F64;
+enum { false = 0, true  = 1, true_overflow, };
+
+#define farray_len(array)                   (SSIZE)sizeof(array) / size_of( typeof((array)[0]))
+#define farray_init(type, ...)              (type[]){__VA_ARGS__}
+
 #define alignas                             _Alignas
 #define alignof                             _Alignof
 #define byte_pad(amount, ...)               Byte glue(_PAD_, __VA_ARGS__) [amount]
-#define farray_len(array)                   (SSIZE)sizeof(array) / size_of( typeof((array)[0]))
-#define farray_init(type, ...)              (type[]){__VA_ARGS__}
 #define def_farray(type, len)               type A ## len ## _ ## type[len]
 #define def_enum(underlying_type, symbol)   underlying_type symbol; enum   symbol
 #define def_struct(symbol)                  struct symbol symbol;   struct symbol
 #define def_union(symbol)                   union  symbol symbol;   union  symbol
 #define def_proc(symbol)                    symbol
 #define opt_args(symbol, ...)               &(symbol){__VA_ARGS__}
-#define ret_type(type)                      type
-#define local_persist                       static
-#define global                              static
-#define offset_of(type, member)             cast(SSIZE, & (((type*) 0)->member))
-#define static_assert                       _Static_assert
-#define typeof                              __typeof__
-#define typeof_ptr(ptr)                     typeof(ptr[0])
-#define typeof_same(a, b)                   _Generic((a), typeof((b)): 1, default: 0)
 
 #define cast(type, data)                    ((type)(data))
 #define pcast(type, data)                   * cast(type*, & (data))
 #define nullptr                             cast(void*, 0)
+#define offset_of(type, member)             cast(SSIZE, & (((type*) 0)->member))
 #define size_of(data)                       cast(SSIZE, sizeof(data))
+
 #define kilo(n)                             (cast(SSIZE, n) << 10)
 #define mega(n)                             (cast(SSIZE, n) << 20)
 #define giga(n)                             (cast(SSIZE, n) << 30)
 #define tera(n)                             (cast(SSIZE, n) << 40)
 
-#define span_iter(type, iter, m_begin, op, m_end)  \
-	tmpl(Iter_Span,type) iter = { \
-		.r = {(m_begin), (m_end)},   \
-		.cursor = (m_begin) };       \
-	iter.cursor op iter.r.end;     \
-	++ iter.cursor
-
-#define def_span(type)                                                 \
-	        def_struct(tmpl(     Span,type)) { type begin; type end; }; \
-	typedef def_struct(tmpl(Iter_Span,type)) { tmpl(Span,type) r; type cursor; }
-
-typedef def_span(S32);
-typedef def_span(U32);
-typedef def_span(SSIZE);
-
 typedef void def_proc(VoidFn) (void);
 #pragma endregion DSL
 
+#pragma region Strings
+typedef unsigned char UTF8;
+typedef def_struct(Str8)       { UTF8* ptr; SSIZE len; }; typedef Str8 Slice_UTF8;
+typedef def_struct(Slice_Str8) { Str8* ptr; SSIZE len; };
+#define lit(string_literal)    (Str8){ (UTF8*) string_literal, size_of(string_literal) - 1 }
+#pragma endregion Strings
+
 #pragma region Debug
 #if !defined(BUILD_DEBUG)
 #define debug_trap()
@@ -133,6 +117,25 @@ void* memory_copy            (void* restrict dest, void const* restrict src, USI
 void* memory_copy_overlapping(void* restrict dest, void const* restrict src, USIZE length);
 B32   memory_zero            (void* dest, USIZE length);
 
+#define check_nil(nil, p) ((p) == 0 || (p) == nil)
+#define set_nil(nil, p)   ((p) = nil)
+
+#define sll_stack_push_n(f, n, next) do { (n)->next = (f); (f) = (n); } while(0)
+
+#define sll_queue_push_nz(nil, f, l, n, next) \
+(                           \
+	check_nil(nil, f) ? (     \
+		(f) = (l) = (n),        \
+		set_nil(nil, (n)->next) \
+	)                         \
+	: (                       \
+		(l)->next=(n),          \
+		(l) = (n),              \
+		set_nil(nil,(n)->next)  \
+	)                         \
+)
+#define sll_queue_push_n(f, l, n, next) sll_queue_push_nz(0, f, l, n, next)
+
 #define def_Slice(type)        \
 def_struct(tmpl(Slice,type)) { \
 	type* ptr; \
@@ -164,24 +167,20 @@ void slice__zero(Slice_Byte mem, SSIZE typewidth);
 	.len = farray_len( farray_init(type, __VA_ARGS__)) \
 }
 
-#define check_nil(nil, p) ((p) == 0 || (p) == nil)
-#define set_nil(nil, p)   ((p) = nil)
+#define span_iter(type, iter, m_begin, op, m_end) \
+	tmpl(Iter_Span,type) iter = { \
+		.r = {(m_begin), (m_end)},  \
+		.cursor = (m_begin) };      \
+	iter.cursor op iter.r.end;    \
+	++ iter.cursor
 
-#define sll_stack_push_n(f, n, next) do { (n)->next = (f); (f) = (n); } while(0)
+#define def_span(type)                                                \
+	        def_struct(tmpl(     Span,type)) { type begin; type end; }; \
+	typedef def_struct(tmpl(Iter_Span,type)) { tmpl(Span,type) r; type cursor; }
 
-#define sll_queue_push_nz(nil, f, l, n, next) \
-(                           \
-	check_nil(nil, f) ? (     \
-		(f) = (l) = (n),        \
-		set_nil(nil, (n)->next) \
-	)                         \
-	: (                       \
-		(l)->next=(n),          \
-		(l) = (n),              \
-		set_nil(nil,(n)->next)  \
-	)                         \
-)
-#define sll_queue_push_n(f, l, n, next) sll_queue_push_nz(0, f, l, n, next)
+typedef def_span(S32);
+typedef def_span(U32);
+typedef def_span(SSIZE);
 #pragma endregion Memory
 
 #pragma region Math
@@ -190,14 +189,6 @@ void slice__zero(Slice_Byte mem, SSIZE typewidth);
 #define clamp_bot(X, B) max(X, B)
 #pragma endregion Math
 
-#pragma region Strings
-typedef unsigned char UTF8;
-typedef def_Slice(UTF8);
-typedef Slice_UTF8 Str8;
-typedef def_Slice(Str8);
-#define lit(string_literal) (Str8){ (UTF8*) string_literal, size_of(string_literal) - 1 }
-#pragma endregion Strings
-
 #pragma region Allocator Interface
 typedef def_enum(U32, AllocatorOp) {
 	AllocatorOp_Alloc_NoZero = 0, // If Alloc exist, so must No_Zero
@@ -604,16 +595,16 @@ void str8gen__append_fmt(Str8Gen* gen, Str8 fmt_template, Slice_A2_Str8* tokens)
 typedef def_struct(FileOpInfo) {
 	Slice_Byte content;
 };
-typedef def_struct(Opts_read_file_contents) {
+typedef def_struct(Opts_file_read_contents) {
 	AllocatorInfo backing;
 	B32           zero_backing;
 	byte_pad(4);
 };
-void api_file_read_contents(FileOpInfo* result, Str8 path, Opts_read_file_contents opts);
+void api_file_read_contents(FileOpInfo* result, Str8 path, Opts_file_read_contents opts);
 void file_write_str8       (Str8 path, Str8 content);
 
-FileOpInfo file__read_contents(Str8 path, Opts_read_file_contents* opts);
-#define file_read_contents(path, ...) file__read_contents(path, &(Opts_read_file_contents){__VA_ARGS__})
+FileOpInfo file__read_contents(Str8 path, Opts_file_read_contents* opts);
+#define file_read_contents(path, ...) file__read_contents(path, opt_args(Opts_file_read_contents, __VA_ARGS__))
 #pragma endregion File System
 
 #pragma region WATL
@@ -655,7 +646,7 @@ typedef def_struct(Opts_watl_lex) {
 };
 void         api_watl_lex(WATL_LexInfo* info, Str8 source, Opts_watl_lex* opts);
 WATL_LexInfo watl__lex   (                    Str8 source, Opts_watl_lex* opts);
-#define watl_lex(source, ...) watl__lex(source, &(Opts_watl_lex){__VA_ARGS__})
+#define watl_lex(source, ...) watl__lex(source, opt_args(Opts_watl_lex, __VA_ARGS__))
 
 typedef Str8 WATL_Node;
 typedef def_Slice(WATL_Node);
@@ -687,7 +678,7 @@ typedef def_struct(Opts_watl_parse) {
 };
 void           api_watl_parse(WATL_ParseInfo* info, Slice_WATL_Tok tokens, Opts_watl_parse* opts);
 WATL_ParseInfo watl__parse   (                      Slice_WATL_Tok tokens, Opts_watl_parse* opts);
-#define watl_parse(tokens, ...) watl__parse(tokens, &(Opts_watl_parse){__VA_ARGS__})
+#define watl_parse(tokens, ...) watl__parse(tokens, opt_args(Opts_watl_parse, __VA_ARGS__))
 
 Str8 watl_dump_listing(AllocatorInfo buffer, Slice_WATL_Line lines);
 #pragma endregion WATL
@@ -1398,10 +1389,8 @@ void kt1cx_init(KT1CX_Info info, KT1CX_InfoMeta m, KT1CX_Byte* result) {
 	assert(m.cell_pool_size >= kilo(4));
 	assert(m.table_size     >= kilo(4));
 	assert(m.type_width     >  0);
-	result->table     = mem_alloc(info.backing_table, m.table_size * m.cell_size);
-	slice_assert(result->table);
-	result->cell_pool = mem_alloc(info.backing_cells, m.cell_size  * m.cell_pool_size);
-	slice_assert(result->cell_pool);
+	result->table     = mem_alloc(info.backing_table, m.table_size * m.cell_size);      slice_assert(result->table);
+	result->cell_pool = mem_alloc(info.backing_cells, m.cell_size  * m.cell_pool_size); slice_assert(result->cell_pool);
 	result->table.len = m.table_size; // Setting to the table number of elements instead of byte length.
 }
 void kt1cx_clear(KT1CX_Byte kt, KT1CX_ByteMeta m) {
@@ -1590,38 +1579,38 @@ Str8 str8__fmt_kt1l(AllocatorInfo ainfo, Slice_Byte* _buffer, KT1L_Str8 table, S
 	UTF8* cursor_buffer    = buffer.ptr;
 	SSIZE buffer_remaining = buffer.len;
 
-	char  curr_code  = * fmt_template.ptr;
 	UTF8* cursor_fmt = fmt_template.ptr;
 	SSIZE left_fmt   = fmt_template.len;
 	while (left_fmt && buffer_remaining)
 	{
+		SSIZE copy_offset = 0;
 		// Forward until we hit the delimiter '<' or the template's contents are exhausted.
-		while (curr_code && curr_code != '<' && cursor_fmt != slice_end(fmt_template)) {
-			*  cursor_buffer = * cursor_fmt;
-			++ cursor_buffer;
-			++ cursor_fmt;
-			-- buffer_remaining;
-			-- left_fmt;
-			curr_code = * cursor_fmt;
+		while (cursor_fmt[copy_offset] != cast(UTF8, '<') && (cursor_fmt + copy_offset) != slice_end(fmt_template)) {
+			++ copy_offset;
 		}
-		if (curr_code == '<')
+		memory_copy(cursor_buffer, cursor_fmt, copy_offset);
+		buffer_remaining -= copy_offset;
+		left_fmt         -= copy_offset;
+		cursor_buffer    += copy_offset;
+		cursor_fmt       += copy_offset;
+
+		if (cursor_fmt[0] == '<')
 		{
-			UTF8* cursor_potential_token = cursor_fmt + 1;
-			SSIZE potential_token_length = 0;
+			UTF8* potential_token_cursor = cursor_fmt + 1;
+			SSIZE potential_token_len    = 0;
 			B32   fmt_overflow = false;
 			for (;;) {
-				UTF8* cursor         = cursor_potential_token + potential_token_length;
+				UTF8* cursor         = potential_token_cursor + potential_token_len;
 				fmt_overflow         = cursor >= slice_end(fmt_template);
-				B32 found_terminator = * (cursor_potential_token + potential_token_length) == '>';
+				B32 found_terminator = * (potential_token_cursor + potential_token_len) == '>';
 				if (fmt_overflow || found_terminator) { break; }
-				++ potential_token_length;
+				++ potential_token_len;
 			}
 			if (fmt_overflow) continue;
 			// Hashing the potential token and cross checking it with our token table
-			U64   key   = 0; hash64_djb8(& key, (Slice_Byte){ cast(Byte*, cursor_potential_token), potential_token_length});
+			U64   key   = 0; hash64_djb8(& key, (Slice_Byte){ cast(Byte*, potential_token_cursor), potential_token_len});
 			Str8* value = nullptr;
-			for (slice_iter(table, token))
-			{
+			for (slice_iter(table, token)) {
 				// We do a linear iteration instead of a hash table lookup because the user should be never substiuting with more than 100 unqiue tokens..
 				if (token->key == key) {
 					value = & token->value;
@@ -1631,32 +1620,26 @@ Str8 str8__fmt_kt1l(AllocatorInfo ainfo, Slice_Byte* _buffer, KT1L_Str8 table, S
 			if (value)
 			{
 				// We're going to appending the string, make sure we have enough space in our buffer.
-				if (ainfo.proc != nullptr && (buffer_remaining - potential_token_length) <= 0) {
-					buffer            = mem_grow(ainfo, buffer, buffer.len + potential_token_length);
-					buffer_remaining += potential_token_length;
-				}
-				SSIZE left         = value->len;
-				U8*   cursor_value = value->ptr;
-
-				while (left && buffer_remaining) {
-					*  cursor_buffer = * cursor_value;
-					++ cursor_buffer;
-					++ cursor_value;
-					-- buffer_remaining;
-					-- left;
+				if (ainfo.proc != nullptr && (buffer_remaining - potential_token_len) <= 0) {
+					buffer            = mem_grow(ainfo, buffer, buffer.len + potential_token_len);
+					buffer_remaining += potential_token_len;
 				}
+				assert((buffer_remaining - potential_token_len) > 0);
+				memory_copy(cursor_buffer, value->ptr, value->len);
 				// Sync cursor format to after the processed token
-				cursor_fmt = cursor_potential_token + potential_token_length + 1;
-				curr_code  = * cursor_fmt;
-				left_fmt  -= potential_token_length + 2; // The 2 here are the '<' & '>' delimiters being omitted.
+				cursor_buffer    += value->len;
+				buffer_remaining -= value->len;
+				cursor_fmt        = potential_token_cursor + potential_token_len + 1;
+				left_fmt         -= potential_token_len + 2; // The 2 here are the '<' & '>' delimiters being omitted.
 				continue;
 			}
+			// If not a value, we do a single copy for the '<' and continue.
 			*  cursor_buffer = * cursor_fmt;
 			++ cursor_buffer;
 			++ cursor_fmt;
 			-- buffer_remaining;
 			-- left_fmt;
-			curr_code = * cursor_fmt;
+			continue;
 		}
 	}
 	* _buffer = buffer;
@@ -1838,12 +1821,12 @@ __declspec(dllimport) MS_BOOL __stdcall GetFileSizeEx(MS_HANDLE hFile, MS_LARGE_
 __declspec(dllimport) MS_DWORD __stdcall GetLastError(void);
 
 inline
-FileOpInfo file__read_contents(Str8 path, Opts_read_file_contents* opts) {
+FileOpInfo file__read_contents(Str8 path, Opts_file_read_contents* opts) {
 	assert(opts != nullptr);
 	FileOpInfo result = {0}; api_file_read_contents(& result, path, * opts);
 	return result;
 }
-void api_file_read_contents(FileOpInfo* result, Str8 path, Opts_read_file_contents opts)
+void api_file_read_contents(FileOpInfo* result, Str8 path, Opts_file_read_contents opts)
 {
 	assert(result != nullptr);
 	slice_assert(path);

Readme.md

@@ -1,12 +1,9 @@
 # WATL Exercise
 
 An exercise on making the "Whitespace Aware Text Layout" parser with different languages or conventions. It simply gets a structural idea of the lines and chunks (visbile and whitespace) for a text file.  
-The purpose of this exercise is convey succiently many core pragmatic concepts for code in a small program.  
 The intent was to use this as a working set of samples for my code visualizing and editing prototyping.
 
-The code conveys a convention for doing "systems" programming I've synthesized after studying how several people in the "handmade" community have written their exposed libraries or codebases.
-
-The goal of the exercise is always the following:
+The goal of is always the following:
 
 ```odin
 start:
@@ -19,6 +16,8 @@ start:
 end
 ```
 
+The exercise is done in several conventions but are generally pragmatic and relatively low abstraction. Favoring composition to construct anything. It succiently conveys many core pragmatic concepts for code in a small source file.  
+
 There are plans for multiple versions of the program:
 
 * V0: Attempt todo a single-threaded example with as little support from the toolchain as possible within the domain of the language. With a single compilation stage.
@@ -33,6 +32,8 @@ Embeddable scripting languages will be embedded as they should be.
 
 ## TODOs
 
+* Fix large-pages not working (at least on my system).
+
 * [x] Single-threaded C example
 * [] Multi-threaded C example
 * [] Add basic timing benchmark to C examples

scripts/build.c_lottes_hybrid.ps1

@@ -131,7 +131,7 @@ $compiler_args += $flag_link_win_rt_static_debug
 # Include setup
 $compiler_args += ($flag_include + $path_root)
 
-$unit_name   = "watl.v0.lottes"
+$unit_name   = "watl.v0.llvm.lottes_hybrid"
 
 # Specify unit to compile
 $unit           = join-path $path_root "C\$unit_name.c"