Merge branch 'daurnimator-fmt-sentinel-pointers'

closes #3972

Author: andrewrk

lib/std/builtin.zig

@@ -185,6 +185,7 @@ pub const TypeInfo = union(enum) {
         child: type,
         is_allowzero: bool,
 
+        /// This field is an optional type.
         /// The type of the sentinel is the element type of the pointer, which is
         /// the value of the `child` field in this struct. However there is no way
         /// to refer to that type here, so we use `var`.
@@ -206,6 +207,7 @@ pub const TypeInfo = union(enum) {
         len: comptime_int,
         child: type,
 
+        /// This field is an optional type.
         /// The type of the sentinel is the element type of the array, which is
         /// the value of the `child` field in this struct. However there is no way
         /// to refer to that type here, so we use `var`.

lib/std/cstr.zig

@@ -28,7 +28,7 @@ test "cstr fns" {
 
 fn testCStrFnsImpl() void {
     testing.expect(cmp("aoeu", "aoez") == -1);
-    testing.expect(mem.len(u8, "123456789") == 9);
+    testing.expect(mem.len("123456789") == 9);
 }
 
 /// Returns a mutable, null-terminated slice with the same length as `slice`.

lib/std/fmt.zig

@@ -441,10 +441,12 @@ pub fn formatType(
                 else => return format(context, Errors, output, "{}@{x}", .{ @typeName(T.Child), @ptrToInt(value) }),
             },
             .Many, .C => {
+                if (ptr_info.sentinel) |sentinel| {
+                    return formatType(mem.span(value), fmt, options, context, Errors, output, max_depth);
+                }
                 if (ptr_info.child == u8) {
                     if (fmt.len > 0 and fmt[0] == 's') {
-                        const len = mem.len(u8, value);
-                        return formatText(value[0..len], fmt, options, context, Errors, output);
+                        return formatText(mem.span(value), fmt, options, context, Errors, output);
                     }
                 }
                 return format(context, Errors, output, "{}@{x}", .{ @typeName(T.Child), @ptrToInt(value) });

lib/std/mem.zig

@@ -333,8 +333,20 @@ pub fn zeroes(comptime T: type) T {
             }
             return array;
         },
-        .Vector, .ErrorUnion, .ErrorSet, .Union, .Fn, .BoundFn, .Type, .NoReturn, .Undefined, .Opaque, .Frame, .AnyFrame,  => {
-            @compileError("Can't set a "++ @typeName(T) ++" to zero.");
+        .Vector,
+        .ErrorUnion,
+        .ErrorSet,
+        .Union,
+        .Fn,
+        .BoundFn,
+        .Type,
+        .NoReturn,
+        .Undefined,
+        .Opaque,
+        .Frame,
+        .AnyFrame,
+        => {
+            @compileError("Can't set a " ++ @typeName(T) ++ " to zero.");
         },
     }
 }
@@ -470,18 +482,115 @@ pub fn eql(comptime T: type, a: []const T, b: []const T) bool {
     return true;
 }
 
-pub fn len(comptime T: type, ptr: [*:0]const T) usize {
-    var count: usize = 0;
-    while (ptr[count] != 0) : (count += 1) {}
-    return count;
-}
-
+/// Deprecated. Use `span`.
 pub fn toSliceConst(comptime T: type, ptr: [*:0]const T) [:0]const T {
-    return ptr[0..len(T, ptr) :0];
+    return ptr[0..len(ptr) :0];
 }
 
+/// Deprecated. Use `span`.
 pub fn toSlice(comptime T: type, ptr: [*:0]T) [:0]T {
-    return ptr[0..len(T, ptr) :0];
+    return ptr[0..len(ptr) :0];
+}
+
+/// Takes a pointer to an array, a sentinel-terminated pointer, or a slice, and
+/// returns a slice. If there is a sentinel on the input type, there will be a
+/// sentinel on the output type. The constness of the output type matches
+/// the constness of the input type. `[*c]` pointers are assumed to be 0-terminated,
+/// and assumed to not allow null.
+pub fn Span(comptime T: type) type {
+    var ptr_info = @typeInfo(T).Pointer;
+    switch (ptr_info.size) {
+        .One => switch (@typeInfo(ptr_info.child)) {
+            .Array => |info| {
+                ptr_info.child = info.child;
+                ptr_info.sentinel = info.sentinel;
+            },
+            else => @compileError("invalid type given to std.mem.Span"),
+        },
+        .C => {
+            ptr_info.sentinel = 0;
+            ptr_info.is_allowzero = false;
+        },
+        .Many, .Slice => {},
+    }
+    ptr_info.size = .Slice;
+    return @Type(std.builtin.TypeInfo{ .Pointer = ptr_info });
+}
+
+test "Span" {
+    testing.expect(Span(*[5]u16) == []u16);
+    testing.expect(Span(*const [5]u16) == []const u16);
+    testing.expect(Span([]u16) == []u16);
+    testing.expect(Span([]const u8) == []const u8);
+    testing.expect(Span([:1]u16) == [:1]u16);
+    testing.expect(Span([:1]const u8) == [:1]const u8);
+    testing.expect(Span([*:1]u16) == [:1]u16);
+    testing.expect(Span([*:1]const u8) == [:1]const u8);
+    testing.expect(Span([*c]u16) == [:0]u16);
+    testing.expect(Span([*c]const u8) == [:0]const u8);
+}
+
+/// Takes a pointer to an array, a sentinel-terminated pointer, or a slice, and
+/// returns a slice. If there is a sentinel on the input type, there will be a
+/// sentinel on the output type. The constness of the output type matches
+/// the constness of the input type.
+pub fn span(ptr: var) Span(@TypeOf(ptr)) {
+    const Result = Span(@TypeOf(ptr));
+    const l = len(ptr);
+    if (@typeInfo(Result).Pointer.sentinel) |s| {
+        return ptr[0..l :s];
+    } else {
+        return ptr[0..l];
+    }
+}
+
+test "span" {
+    var array: [5]u16 = [_]u16{ 1, 2, 3, 4, 5 };
+    const ptr = array[0..2 :3].ptr;
+    testing.expect(eql(u16, span(ptr), &[_]u16{ 1, 2 }));
+    testing.expect(eql(u16, span(&array), &[_]u16{ 1, 2, 3, 4, 5 }));
+}
+
+/// Takes a pointer to an array, an array, a sentinel-terminated pointer,
+/// or a slice, and returns the length.
+pub fn len(ptr: var) usize {
+    return switch (@typeInfo(@TypeOf(ptr))) {
+        .Array => |info| info.len,
+        .Pointer => |info| switch (info.size) {
+            .One => switch (@typeInfo(info.child)) {
+                .Array => |x| x.len,
+                else => @compileError("invalid type given to std.mem.length"),
+            },
+            .Many => if (info.sentinel) |sentinel|
+                indexOfSentinel(info.child, sentinel, ptr)
+            else
+                @compileError("length of pointer with no sentinel"),
+            .C => indexOfSentinel(info.child, 0, ptr),
+            .Slice => ptr.len,
+        },
+        else => @compileError("invalid type given to std.mem.length"),
+    };
+}
+
+test "len" {
+    testing.expect(len("aoeu") == 4);
+
+    {
+        var array: [5]u16 = [_]u16{ 1, 2, 3, 4, 5 };
+        testing.expect(len(&array) == 5);
+        testing.expect(len(array[0..3]) == 3);
+        array[2] = 0;
+        const ptr = array[0..2 :0].ptr;
+        testing.expect(len(ptr) == 2);
+    }
+}
+
+pub fn indexOfSentinel(comptime Elem: type, comptime sentinel: Elem, ptr: [*:sentinel]const Elem) usize {
+    var i: usize = 0;
+    while (ptr[i] != sentinel) {
+        i += 1;
+    }
+    return i;
 }
 
 /// Returns true if all elements in a slice are equal to the scalar value provided
@@ -1753,8 +1862,13 @@ fn SubArrayPtrReturnType(comptime T: type, comptime length: usize) type {
     return *[length]meta.Child(meta.Child(T));
 }
 
-///Given a pointer to an array, returns a pointer to a portion of that array, preserving constness.
-pub fn subArrayPtr(ptr: var, comptime start: usize, comptime length: usize) SubArrayPtrReturnType(@TypeOf(ptr), length) {
+/// Given a pointer to an array, returns a pointer to a portion of that array, preserving constness.
+/// TODO this will be obsoleted by https://github.com/ziglang/zig/issues/863
+pub fn subArrayPtr(
+    ptr: var,
+    comptime start: usize,
+    comptime length: usize,
+) SubArrayPtrReturnType(@TypeOf(ptr), length) {
     assert(start + length <= ptr.*.len);
 
     const ReturnType = SubArrayPtrReturnType(@TypeOf(ptr), length);

lib/std/meta.zig

@@ -115,6 +115,32 @@ test "std.meta.Child" {
     testing.expect(Child(?u8) == u8);
 }
 
+/// Given a type with a sentinel e.g. `[:0]u8`, returns the sentinel
+pub fn Sentinel(comptime T: type) Child(T) {
+    // comptime asserts that ptr has a sentinel
+    switch (@typeInfo(T)) {
+        .Array => |arrayInfo| {
+            return comptime arrayInfo.sentinel.?;
+        },
+        .Pointer => |ptrInfo| {
+            switch (ptrInfo.size) {
+                .Many, .Slice => {
+                    return comptime ptrInfo.sentinel.?;
+                },
+                else => {},
+            }
+        },
+        else => {},
+    }
+    @compileError("not a sentinel type, found '" ++ @typeName(T) ++ "'");
+}
+
+test "std.meta.Sentinel" {
+    testing.expectEqual(@as(u8, 0), Sentinel([:0]u8));
+    testing.expectEqual(@as(u8, 0), Sentinel([*:0]u8));
+    testing.expectEqual(@as(u8, 0), Sentinel([5:0]u8));
+}
+
 pub fn containerLayout(comptime T: type) TypeInfo.ContainerLayout {
     return switch (@typeInfo(T)) {
         .Struct => |info| info.layout,

lib/std/net.zig

@@ -352,7 +352,7 @@ pub const Address = extern union {
                     unreachable;
                 }
 
-                const path_len = std.mem.len(u8, @ptrCast([*:0]const u8, &self.un.path));
+                const path_len = std.mem.len(@ptrCast([*:0]const u8, &self.un.path));
                 return @intCast(os.socklen_t, @sizeOf(os.sockaddr_un) - self.un.path.len + path_len);
             },
             else => unreachable,

lib/std/os.zig

@@ -1095,7 +1095,7 @@ pub fn createNullDelimitedEnvMap(allocator: *mem.Allocator, env_map: *const std.
 
 pub fn freeNullDelimitedEnvMap(allocator: *mem.Allocator, envp_buf: []?[*:0]u8) void {
     for (envp_buf) |env| {
-        const env_buf = if (env) |ptr| ptr[0 .. mem.len(u8, ptr) + 1] else break;
+        const env_buf = if (env) |ptr| ptr[0 .. mem.len(ptr) + 1] else break;
         allocator.free(env_buf);
     }
     allocator.free(envp_buf);

lib/std/special/c.zig

@@ -47,7 +47,7 @@ fn strcmp(s1: [*:0]const u8, s2: [*:0]const u8) callconv(.C) c_int {
 }
 
 fn strlen(s: [*:0]const u8) callconv(.C) usize {
-    return std.mem.len(u8, s);
+    return std.mem.len(s);
 }
 
 fn strncmp(_l: [*:0]const u8, _r: [*:0]const u8, _n: usize) callconv(.C) c_int {

src-self-hosted/main.zig

@@ -792,7 +792,7 @@ async fn fmtPath(fmt: *Fmt, file_path_ref: []const u8, check_mode: bool) FmtErro
 }
 
 fn cmdVersion(allocator: *Allocator, args: []const []const u8) !void {
-    try stdout.print("{}\n", .{std.mem.toSliceConst(u8, c.ZIG_VERSION_STRING)});
+    try stdout.print("{}\n", .{c.ZIG_VERSION_STRING});
 }
 
 fn cmdHelp(allocator: *Allocator, args: []const []const u8) !void {
@@ -863,12 +863,12 @@ fn cmdInternalBuildInfo(allocator: *Allocator, args: []const []const u8) !void {
         \\ZIG_DIA_GUIDS_LIB    {}
         \\
     , .{
-        std.mem.toSliceConst(u8, c.ZIG_CMAKE_BINARY_DIR),
-        std.mem.toSliceConst(u8, c.ZIG_CXX_COMPILER),
-        std.mem.toSliceConst(u8, c.ZIG_LLD_INCLUDE_PATH),
-        std.mem.toSliceConst(u8, c.ZIG_LLD_LIBRARIES),
-        std.mem.toSliceConst(u8, c.ZIG_LLVM_CONFIG_EXE),
-        std.mem.toSliceConst(u8, c.ZIG_DIA_GUIDS_LIB),
+        c.ZIG_CMAKE_BINARY_DIR,
+        c.ZIG_CXX_COMPILER,
+        c.ZIG_LLD_INCLUDE_PATH,
+        c.ZIG_LLD_LIBRARIES,
+        c.ZIG_LLVM_CONFIG_EXE,
+        c.ZIG_DIA_GUIDS_LIB,
     });
 }
 

src-self-hosted/translate_c.zig

@@ -4849,7 +4849,7 @@ fn transPreprocessorEntities(c: *Context, unit: *ZigClangASTUnit) Error!void {
                 }
 
                 const begin_c = ZigClangSourceManager_getCharacterData(c.source_manager, begin_loc);
-                const slice = begin_c[0..mem.len(u8, begin_c)];
+                const slice = begin_c[0..mem.len(begin_c)];
 
                 tok_list.shrink(0);
                 var tokenizer = std.c.Tokenizer{

test/stage1/behavior/misc.zig

@@ -335,7 +335,7 @@ test "string concatenation" {
     comptime expect(@TypeOf(a) == *const [12:0]u8);
     comptime expect(@TypeOf(b) == *const [12:0]u8);
 
-    const len = mem.len(u8, b);
+    const len = mem.len(b);
     const len_with_null = len + 1;
     {
         var i: u32 = 0;