Remove number casting syntax

doc/langref.html.in

@@ -1355,7 +1355,7 @@ var some_integers: [100]i32 = undefined;
 
 test "modify an array" {
     for (some_integers) |*item, i| {
-        item.* = i32(i);
+        item.* = @intCast(i32, i);
     }
     assert(some_integers[10] == 10);
     assert(some_integers[99] == 99);
@@ -1397,8 +1397,8 @@ var fancy_array = init: {
     var initial_value: [10]Point = undefined;
     for (initial_value) |*pt, i| {
         pt.* = Point{
-            .x = i32(i),
-            .y = i32(i) * 2,
+            .x = @intCast(i32, i),
+            .y = @intCast(i32, i) * 2,
         };
     }
     break :init initial_value;
@@ -2410,7 +2410,7 @@ test "for basics" {
     var sum2: i32 = 0;
     for (items) |value, i| {
         assert(@typeOf(i) == usize);
-        sum2 += i32(i);
+        sum2 += @intCast(i32, i);
     }
     assert(sum2 == 10);
 }
@@ -5730,7 +5730,7 @@ comptime {
       {#code_begin|test_err|attempt to cast negative value to unsigned integer#}
 comptime {
     const value: i32 = -1;
-    const unsigned = u32(value);
+    const unsigned = @intCast(u32, value);
 }
       {#code_end#}
       <p>At runtime crashes with the message <code>attempt to cast negative value to unsigned integer</code> and a stack trace.</p>
@@ -5744,7 +5744,7 @@ comptime {
       {#code_begin|test_err|cast from 'u16' to 'u8' truncates bits#}
 comptime {
     const spartan_count: u16 = 300;
-    const byte = u8(spartan_count);
+    const byte = @intCast(u8, spartan_count);
 }
       {#code_end#}
       <p>At runtime crashes with the message <code>integer cast truncated bits</code> and a stack trace.</p>

example/hello_world/hello_libc.zig

@@ -8,7 +8,7 @@ const c = @cImport({
 const msg = c"Hello, world!\n";
 
 export fn main(argc: c_int, argv: **u8) c_int {
-    if (c.printf(msg) != c_int(c.strlen(msg))) return -1;
+    if (c.printf(msg) != @intCast(c_int, c.strlen(msg))) return -1;
 
     return 0;
 }

src/all_types.hpp

@@ -1357,6 +1357,10 @@ enum BuiltinFnId {
     BuiltinFnIdMod,
     BuiltinFnIdSqrt,
     BuiltinFnIdTruncate,
+    BuiltinFnIdIntCast,
+    BuiltinFnIdFloatCast,
+    BuiltinFnIdIntToFloat,
+    BuiltinFnIdFloatToInt,
     BuiltinFnIdIntType,
     BuiltinFnIdSetCold,
     BuiltinFnIdSetRuntimeSafety,
@@ -2040,6 +2044,10 @@ enum IrInstructionId {
     IrInstructionIdCmpxchg,
     IrInstructionIdFence,
     IrInstructionIdTruncate,
+    IrInstructionIdIntCast,
+    IrInstructionIdFloatCast,
+    IrInstructionIdIntToFloat,
+    IrInstructionIdFloatToInt,
     IrInstructionIdIntType,
     IrInstructionIdBoolNot,
     IrInstructionIdMemset,
@@ -2632,6 +2640,34 @@ struct IrInstructionTruncate {
     IrInstruction *target;
 };
 
+struct IrInstructionIntCast {
+    IrInstruction base;
+
+    IrInstruction *dest_type;
+    IrInstruction *target;
+};
+
+struct IrInstructionFloatCast {
+    IrInstruction base;
+
+    IrInstruction *dest_type;
+    IrInstruction *target;
+};
+
+struct IrInstructionIntToFloat {
+    IrInstruction base;
+
+    IrInstruction *dest_type;
+    IrInstruction *target;
+};
+
+struct IrInstructionFloatToInt {
+    IrInstruction base;
+
+    IrInstruction *dest_type;
+    IrInstruction *target;
+};
+
 struct IrInstructionIntType {
     IrInstruction base;
 

src/codegen.cpp

@@ -4722,6 +4722,10 @@ static LLVMValueRef ir_render_instruction(CodeGen *g, IrExecutable *executable,
         case IrInstructionIdPromiseResultType:
         case IrInstructionIdAwaitBookkeeping:
         case IrInstructionIdAddImplicitReturnType:
+        case IrInstructionIdIntCast:
+        case IrInstructionIdFloatCast:
+        case IrInstructionIdIntToFloat:
+        case IrInstructionIdFloatToInt:
             zig_unreachable();
 
         case IrInstructionIdReturn:
@@ -6310,6 +6314,10 @@ static void define_builtin_fns(CodeGen *g) {
     create_builtin_fn(g, BuiltinFnIdCmpxchgStrong, "cmpxchgStrong", 6);
     create_builtin_fn(g, BuiltinFnIdFence, "fence", 1);
     create_builtin_fn(g, BuiltinFnIdTruncate, "truncate", 2);
+    create_builtin_fn(g, BuiltinFnIdIntCast, "intCast", 2);
+    create_builtin_fn(g, BuiltinFnIdFloatCast, "floatCast", 2);
+    create_builtin_fn(g, BuiltinFnIdIntToFloat, "intToFloat", 2);
+    create_builtin_fn(g, BuiltinFnIdFloatToInt, "floatToInt", 2);
     create_builtin_fn(g, BuiltinFnIdCompileErr, "compileError", 1);
     create_builtin_fn(g, BuiltinFnIdCompileLog, "compileLog", SIZE_MAX);
     create_builtin_fn(g, BuiltinFnIdIntType, "IntType", 2); // TODO rename to Int