stage2 提供词法分析器通过与stage1的差异测试 stage1提供数组到切片的隐式转换

stage2/utils/byte_view.spl

@@ -35,6 +35,14 @@ fn from_slice(bytes: []u8): ByteView {
     ret ByteView { .ptr = bytes.ptr, .len = bytes.len };
 }
 
+// 以 NUL 结尾的 C 字符串转为视图；不复制、不拥有。null 规范化为空视图。
+fn from_cstr(text: *u8): ByteView {
+    if text == null { ret empty(); }
+    var len: usize = 0;
+    while text[len] != 0 { len = len + 1; }
+    ret from_slice(.{ .ptr = text, .len = len });
+}
+
 fn as_slice(self: ByteView): []u8 {
     ret .{ .ptr = self.ptr, .len = self.len };
 }
@@ -213,6 +221,14 @@ fn test_construction_and_access() {
 
     var bytes = as_slice(text);
     @assert(bytes.ptr == text.ptr && bytes.len == text.len, "slice conversion borrows");
+
+    var cstr_data: [4]u8 = .[104, 105, 0, 88];
+    var cstr = from_cstr(&cstr_data[0]);
+    @assert(cstr.len == 2 && cstr.ptr[0] == 104 && cstr.ptr[1] == 105,
+            "from_cstr stops at NUL");
+    @assert(from_cstr(&cstr_data[2]).len == 0, "from_cstr at NUL is empty");
+    var none = from_cstr(null);
+    @assert(none.ptr == null && none.len == 0, "from_cstr null normalized");
 }
 
 fn test_compare_and_slice() {

stage2/utils/fmt.spl

@@ -5,6 +5,8 @@
 
 type .{ .Error, .Status, .make } = @import("error.spl");
 type .{ .ByteBuffer } = @import("byte_buffer.spl");
+type .{ .ByteView } = @import("byte_view.spl");
+type .{ .Value, .int, .uint, .char, .str, .addr } = @import("value.spl");
 type .{ .BigInt, .Bits } = @import("mp.spl");
 
 // 注意：radix 边界直接用字面量。stage1 刻意不做常量求值/折叠，
@@ -69,6 +71,158 @@ fn append_bool(out: *ByteBuffer, value: bool): Status {
     ret ByteBuffer.append(out, "false");
 }
 
+// ---- 显式装箱值格式化 ----
+
+// Int/Uint 落到整数；signed 仅对 Int 的十进制生效，其余按无符号位模式输出。
+fn append_value_int(out: *ByteBuffer, value: Value, radix: u8, uppercase: bool,
+                    signed: bool): Status {
+    match value {
+        .Int(v) => {
+            if signed && radix == 10 { ret append_i64(out, v, radix, uppercase); }
+            ret append_u64(out, v as u64, radix, uppercase);
+        },
+        .Uint(v) => { ret append_u64(out, v, radix, uppercase); },
+        _ => { ret .Err(make(.InvalidFormat)); },
+    }
+}
+
+fn append_value_char(out: *ByteBuffer, value: Value): Status {
+    match value {
+        .Char(c) => { ret ByteBuffer.append_byte(out, c); },
+        _ => { ret .Err(make(.InvalidFormat)); },
+    }
+}
+
+fn append_value_str(out: *ByteBuffer, value: Value): Status {
+    match value {
+        .Str(view) => { ret ByteBuffer.append(out, ByteView.as_slice(view)); },
+        _ => { ret .Err(make(.InvalidFormat)); },
+    }
+}
+
+fn append_value_ptr(out: *ByteBuffer, value: Value): Status {
+    match value {
+        .Ptr(p) => {
+            match ByteBuffer.append(out, "0x") {
+                .Ok => {},
+                .Err(error) => { ret .Err(error); },
+            }
+            ret append_u64(out, p as usize, 16, false);
+        },
+        _ => { ret .Err(make(.InvalidFormat)); },
+    }
+}
+
+// {} 自动格式化：按 Value 变体选择默认表示。
+fn append_value_auto(out: *ByteBuffer, value: Value): Status {
+    match value {
+        .Int(v) => { ret append_i64(out, v, 10, false); },
+        .Uint(v) => { ret append_u64(out, v, 10, false); },
+        .Bool(v) => { ret append_bool(out, v); },
+        .Char(c) => { ret ByteBuffer.append_byte(out, c); },
+        .Str(view) => { ret ByteBuffer.append(out, ByteView.as_slice(view)); },
+        .Ptr(p) => {
+            match ByteBuffer.append(out, "0x") {
+                .Ok => {},
+                .Err(error) => { ret .Err(error); },
+            }
+            ret append_u64(out, p as usize, 16, false);
+        },
+    }
+    ret .Ok;
+}
+
+// {} 风格：{} 自动；{:d} {:i} {:u} {:x} {:X} {:o} {:b} {:c} {:s} {:p} 指定。
+// {{ 与 }} 输出字面量括号。参数按序从 args 消费；数量不足、类型不符、未知转换
+// 或括号不配返回 Err(.InvalidFormat)。失败不产生部分输出：先在临时缓冲完成。
+fn format(out: *ByteBuffer, fmt: []u8, args: []Value): Status {
+    var temp = ByteBuffer.new();
+    defer { ByteBuffer.drop(&temp); }
+
+    var argi: usize = 0;
+    var i: usize = 0;
+    while i < fmt.len {
+        var c = fmt.ptr[i];
+
+        if c == '{' {
+            if i + 1 < fmt.len && fmt.ptr[i + 1] == '{' {
+                match ByteBuffer.append_byte(&temp, '{') {
+                    .Ok => {},
+                    .Err(error) => { ret .Err(error); },
+                }
+                i = i + 2;
+                continue;
+            }
+
+            var j = i + 1;
+            if j < fmt.len && fmt.ptr[j] == ':' { j = j + 1; }
+            var has_spec = false;
+            var spec: u8 = 0;
+            if j < fmt.len && fmt.ptr[j] != '}' {
+                has_spec = true;
+                spec = fmt.ptr[j];
+                j = j + 1;
+            }
+            if j >= fmt.len || fmt.ptr[j] != '}' { ret .Err(make(.InvalidFormat)); }
+            i = j + 1;
+
+            if argi >= args.len { ret .Err(make(.InvalidFormat)); }
+            var value = args.ptr[argi];
+            argi = argi + 1;
+
+            var step: Status = .Ok;
+            if !has_spec {
+                step = append_value_auto(&temp, value);
+            } else if spec == 'd' || spec == 'i' {
+                step = append_value_int(&temp, value, 10, false, true);
+            } else if spec == 'u' {
+                step = append_value_int(&temp, value, 10, false, false);
+            } else if spec == 'x' {
+                step = append_value_int(&temp, value, 16, false, false);
+            } else if spec == 'X' {
+                step = append_value_int(&temp, value, 16, true, false);
+            } else if spec == 'o' {
+                step = append_value_int(&temp, value, 8, false, false);
+            } else if spec == 'b' {
+                step = append_value_int(&temp, value, 2, false, false);
+            } else if spec == 'c' {
+                step = append_value_char(&temp, value);
+            } else if spec == 's' {
+                step = append_value_str(&temp, value);
+            } else if spec == 'p' {
+                step = append_value_ptr(&temp, value);
+            } else {
+                ret .Err(make(.InvalidFormat));
+            }
+            match step {
+                .Ok => {},
+                .Err(error) => { ret .Err(error); },
+            }
+            continue;
+        }
+
+        if c == '}' {
+            if i + 1 < fmt.len && fmt.ptr[i + 1] == '}' {
+                match ByteBuffer.append_byte(&temp, '}') {
+                    .Ok => {},
+                    .Err(error) => { ret .Err(error); },
+                }
+                i = i + 2;
+                continue;
+            }
+            ret .Err(make(.InvalidFormat));
+        }
+
+        match ByteBuffer.append_byte(&temp, c) {
+            .Ok => {},
+            .Err(error) => { ret .Err(error); },
+        }
+        i = i + 1;
+    }
+
+    ret ByteBuffer.append(out, ByteBuffer.as_slice(temp));
+}
+
 // ---- 测试辅助 ----
 
 fn expect_ok(result: Status) {
@@ -262,10 +416,96 @@ fn test_composition_and_growth() {
     @assert(s.ptr[1] == ',', "0 then comma");
 }
 
+fn expect_format(fmt: []u8, args: []Value, expected: []u8) {
+    var b = ByteBuffer.new();
+    defer { ByteBuffer.drop(&b); }
+    expect_ok(format(&b, fmt, args));
+    expect_text(&b, expected);
+}
+
+fn expect_invalid_format(fmt: []u8, args: []Value) {
+    var b = ByteBuffer.new();
+    defer { ByteBuffer.drop(&b); }
+    match format(&b, fmt, args) {
+        .Ok => { @assert(false, "expected invalid format"); },
+        .Err(error) => {
+            match error.kind {
+                .InvalidFormat => {},
+                _ => { @assert(false, "expected InvalidFormat kind"); },
+            }
+        },
+    }
+}
+
+fn test_format_specs() {
+    var one: [1]Value = .[int(-5)];
+    expect_format("{}", one[0..1], "-5");
+    expect_format("{d}", one[0..1], "-5");
+    expect_format("{i}", one[0..1], "-5");
+
+    var u: [1]Value = .[uint(7)];
+    expect_format("{}", u[0..1], "7");
+    expect_format("{u}", u[0..1], "7");
+    expect_format("{x}", u[0..1], "7");
+    expect_format("{o}", u[0..1], "7");
+    expect_format("{b}", u[0..1], "111");
+
+    var hex: [1]Value = .[uint(255)];
+    expect_format("{x}", hex[0..1], "ff");
+    expect_format("{X}", hex[0..1], "FF");
+
+    var oct: [1]Value = .[uint(8)];
+    expect_format("{o}", oct[0..1], "10");
+
+    var bits: [1]Value = .[uint(5)];
+    expect_format("{b}", bits[0..1], "101");
+
+    var ch: [1]Value = .[char('A')];
+    expect_format("{}", ch[0..1], "A");
+    expect_format("{c}", ch[0..1], "A");
+
+    var text: [1]Value = .[str("hi")];
+    expect_format("{}", text[0..1], "hi");
+    expect_format("{s}", text[0..1], "hi");
+
+    var nul: [1]Value = .[addr(null)];
+    expect_format("{p}", nul[0..1], "0x0");
+
+    var none: [1]Value = .[int(0)];
+    expect_format("100%", none[0..0], "100%");
+    expect_format("{{}}", none[0..0], "{}");
+    expect_format("plain", none[0..0], "plain");
+}
+
+fn test_format_mixed_and_failure() {
+    var args: [3]Value = .[int(42), str("x"), char('!')];
+    expect_format("n={} s={} c={}", args[0..3], "n=42 s=x c=!");
+    expect_format("n={} s={s} c={c}", args[0..3], "n=42 s=x c=!");
+
+    // 失败不改动输出，且不产生部分结果：前两个转换成功、第三个参数不足。
+    var b = ByteBuffer.new();
+    defer { ByteBuffer.drop(&b); }
+    expect_ok(ByteBuffer.append(&b, "prefix"));
+    var two: [2]Value = .[int(1), int(2)];
+    expect_invalid_format("a={} b={} c={}", two[0..2]);
+    expect_text(&b, "prefix");
+
+    var only_int: [1]Value = .[int(1)];
+    expect_invalid_format("{} {}", only_int[0..1]);
+    expect_invalid_format("{}", only_int[0..0]);
+    expect_invalid_format("trailing {", only_int[0..1]);
+    expect_invalid_format("dangling }", only_int[0..1]);
+    expect_invalid_format("{q}", only_int[0..1]);
+    expect_invalid_format("{s}", only_int[0..1]);
+    expect_invalid_format("{c}", only_int[0..1]);
+}
+
 fn main() {
     test_unsigned();
     test_signed();
     test_usize_bool_and_errors();
     test_mpa_integer_formatting();
     test_composition_and_growth();
+    test_format_specs();
+    test_format_mixed_and_failure();
 }

stage2/utils/log.spl

@@ -0,0 +1,205 @@
+// Expect retcode = 0
+
+// 编译器自身日志（区别于 diag：diag 负责带源码位置的诊断）。
+// 输出到 stderr，格式稳定：[LEVEL] name: message\n。
+// 消息为借用；Logger 不拥有 name。fatal 打印后终止进程（退出码 1）。
+// write 返回 Status；OOM 归类为 .Overflow（无独立内存耗尽错误码）。
+
+type .{ .Status, .make } = @import("error.spl");
+type .{ .ByteView } = @import("byte_view.spl");
+type .{ .ByteBuffer } = @import("byte_buffer.spl");
+type .{ .format } = @import("fmt.spl");
+type .{ .Value, .int, .str } = @import("value.spl");
+type .{ .eprint } = @import("io.spl");
+type Exit = @import("../sys/exit.spl");
+
+type Level = enum { Debug, Info, Warn, Error, Fatal }
+
+fn level_name(level: Level): []u8 {
+    match level {
+        .Debug => { ret "DEBUG"; },
+        .Info => { ret "INFO "; },
+        .Warn => { ret "WARN "; },
+        .Error => { ret "ERROR"; },
+        .Fatal => { ret "FATAL"; },
+    }
+    ret "NOTSET";
+}
+
+type Logger = @self;
+
+name: ByteView,
+min_level: Level,
+
+fn new(name: ByteView): Logger {
+    ret Logger { .name = name, .min_level = .Debug };
+}
+
+fn set_level(self: *Logger, level: Level) {
+    self.min_level = level;
+}
+
+fn enabled(self: Logger, level: Level): bool {
+    ret @enum2int(level) >= @enum2int(self.min_level);
+}
+
+// 追加成功返回 true；失败（仅 OOM）返回 false。
+fn push(line: *ByteBuffer, text: []u8): bool {
+    match ByteBuffer.append(line, text) {
+        .Ok => { ret true; },
+        .Err(error) => { ret false; },
+    }
+    ret false;
+}
+
+// 低于 min_level 的日志不写出，直接成功。
+fn write(self: *Logger, level: Level, message: []u8): Status {
+    if !enabled(self.*, level) { ret .Ok; }
+    var line = ByteBuffer.new();
+    defer { ByteBuffer.drop(&line); }
+    if !push(&line, "[") { ret .Err(make(.Overflow)); }
+    if !push(&line, level_name(level)) { ret .Err(make(.Overflow)); }
+    if !push(&line, "] ") { ret .Err(make(.Overflow)); }
+    if !push(&line, ByteView.as_slice(self.name)) { ret .Err(make(.Overflow)); }
+    if !push(&line, ": ") { ret .Err(make(.Overflow)); }
+    if !push(&line, message) { ret .Err(make(.Overflow)); }
+    if !push(&line, "\n") { ret .Err(make(.Overflow)); }
+    ret eprint(ByteBuffer.as_slice(line));
+}
+
+fn debug(self: *Logger, message: []u8): Status {
+    ret write(self, .Debug, message);
+}
+
+fn info(self: *Logger, message: []u8): Status {
+    ret write(self, .Info, message);
+}
+
+fn warn(self: *Logger, message: []u8): Status {
+    ret write(self, .Warn, message);
+}
+
+fn error(self: *Logger, message: []u8): Status {
+    ret write(self, .Error, message);
+}
+
+// 尽力写出后终止；写失败无法挽回，直接退出。
+fn fatal(self: *Logger, message: []u8) {
+    write(self, .Fatal, message);
+    Exit.fail();
+}
+
+// ---- 格式化入口（显式装箱，见 value.spl / fmt.format）----
+
+// 先格式化到临时缓冲；格式错误在写出前返回 Err(.InvalidFormat)。
+fn logf(self: *Logger, level: Level, fmt: []u8, args: []Value): Status {
+    if !enabled(self.*, level) { ret .Ok; }
+    var message = ByteBuffer.new();
+    defer { ByteBuffer.drop(&message); }
+    match format(&message, fmt, args) {
+        .Ok => {},
+        .Err(error) => { ret .Err(error); },
+    }
+    ret write(self, level, ByteBuffer.as_slice(message));
+}
+
+fn debugf(self: *Logger, fmt: []u8, args: []Value): Status {
+    ret logf(self, .Debug, fmt, args);
+}
+
+fn infof(self: *Logger, fmt: []u8, args: []Value): Status {
+    ret logf(self, .Info, fmt, args);
+}
+
+fn warnf(self: *Logger, fmt: []u8, args: []Value): Status {
+    ret logf(self, .Warn, fmt, args);
+}
+
+fn errorf(self: *Logger, fmt: []u8, args: []Value): Status {
+    ret logf(self, .Error, fmt, args);
+}
+
+// 格式化后终止；格式化失败也要退出（避免吞掉致命错误）。
+fn fatalf(self: *Logger, fmt: []u8, args: []Value) {
+    logf(self, .Fatal, fmt, args);
+    Exit.fail();
+}
+
+// ---- 测试 ----
+
+fn expect_name(level: Level, expected: []u8) {
+    @assert(ByteView.eq(ByteView.from_slice(level_name(level)),
+                        ByteView.from_slice(expected)),
+            "level name");
+}
+
+fn expect_ok(result: Status) {
+    match result {
+        .Ok => {},
+        .Err(error) => { @assert(false, "unexpected log error"); },
+    }
+}
+
+fn test_level_names() {
+    expect_name(.Debug, "DEBUG");
+    expect_name(.Info, "INFO ");
+    expect_name(.Warn, "WARN ");
+    expect_name(.Error, "ERROR");
+    expect_name(.Fatal, "FATAL");
+
+    var debug: Level = .Debug;
+    var info: Level = .Info;
+    var warn: Level = .Warn;
+    var err: Level = .Error;
+    var fatal: Level = .Fatal;
+    @assert(@enum2int(fatal) > @enum2int(err), "fatal above error");
+    @assert(@enum2int(err) > @enum2int(warn), "error above warn");
+    @assert(@enum2int(warn) > @enum2int(info), "warn above info");
+    @assert(@enum2int(info) > @enum2int(debug), "info above debug");
+}
+
+fn test_filter_and_write() {
+    var log = new(ByteView.from_slice("splc1"));
+    @assert(enabled(log, .Debug), "default shows debug");
+    @assert(enabled(log, .Fatal), "default shows fatal");
+
+    set_level(&log, .Warn);
+    @assert(!enabled(log, .Debug), "filtered debug");
+    @assert(!enabled(log, .Info), "filtered info");
+    @assert(enabled(log, .Warn), "warn passes");
+    @assert(enabled(log, .Error), "error passes");
+
+    // 被过滤的写入不产生输出，且返回成功。
+    expect_ok(debug(&log, "hidden"));
+    expect_ok(write(&log, .Info, "hidden too"));
+    expect_ok(error(&log, "visible"));
+}
+
+fn test_logf() {
+    var log = new(ByteView.from_slice("splc1"));
+    set_level(&log, .Error);
+
+    var args: [2]Value = .[int(3), str("x")];
+
+    // 被过滤的级别不写出，直接成功。
+    expect_ok(infof(&log, "n={} s={}", args[0..2]));
+    // 通过级别正常写出。
+    expect_ok(errorf(&log, "n={} s={}", args[0..2]));
+
+    // 格式化失败向上传播，且不产出消息。
+    match errorf(&log, "{} {}", args[0..1]) {
+        .Ok => { @assert(false, "expected invalid format"); },
+        .Err(error) => {
+            match error.kind {
+                .InvalidFormat => {},
+                _ => { @assert(false, "expected InvalidFormat"); },
+            }
+        },
+    }
+}
+
+fn main() {
+    test_level_names();
+    test_filter_and_write();
+    test_logf();
+}

stage2/utils/value.spl

@@ -0,0 +1,100 @@
+// Expect retcode = 0
+
+// 显式装箱值：运行期格式化用的最小 ADT。调用方显式构造（如 Value.int(x)），
+// 库内用 match 分派。不做隐式装箱、不带运行期类型系统。
+// Str 借用 ByteView，不拥有存储；Ptr 是类型擦除指针。
+
+type .{ .ByteView } = @import("byte_view.spl");
+
+type Value = enum {
+    Int: i64,
+    Uint: u64,
+    Bool: bool,
+    Char: u8,
+    Str: ByteView,
+    Ptr: *any,
+}
+
+fn int(value: i64): Value {
+    ret .Int(value);
+}
+
+fn uint(value: u64): Value {
+    ret .Uint(value);
+}
+
+fn boolean(value: bool): Value {
+    ret .Bool(value);
+}
+
+fn char(value: u8): Value {
+    ret .Char(value);
+}
+
+fn str(bytes: []u8): Value {
+    ret .Str(ByteView.from_slice(bytes));
+}
+
+fn str_view(view: ByteView): Value {
+    ret .Str(view);
+}
+
+fn addr(value: *any): Value {
+    ret .Ptr(value);
+}
+
+// ---- 测试 ----
+
+fn test_int_and_uint() {
+    match int(-5) {
+        .Int(v) => { @assert(v == -5, "int payload"); },
+        _ => { @assert(false, "expected int"); },
+    }
+    match uint(5) {
+        .Uint(v) => { @assert(v == 5, "uint payload"); },
+        _ => { @assert(false, "expected uint"); },
+    }
+}
+
+fn test_scalars() {
+    match boolean(true) {
+        .Bool(v) => { @assert(v, "bool payload"); },
+        _ => { @assert(false, "expected bool"); },
+    }
+    match char('A') {
+        .Char(v) => { @assert(v == 'A', "char payload"); },
+        _ => { @assert(false, "expected char"); },
+    }
+}
+
+fn test_borrowed_str() {
+    var text: [3]u8 = .[104, 105, 0];
+    var s = str(text[0..2]);
+    match s {
+        .Str(view) => {
+            @assert(view.len == 2 && view.ptr[0] == 104 && view.ptr[1] == 105,
+                    "str payload borrows");
+        },
+        _ => { @assert(false, "expected str"); },
+    }
+    @assert(text[0] == 104, "source untouched");
+}
+
+fn test_ptr() {
+    var x: i32 = 7;
+    match addr(&x as *any) {
+        .Ptr(p) => { @assert((p as *i32).* == 7, "ptr payload"); },
+        _ => { @assert(false, "expected ptr"); },
+    }
+    match addr(null) {
+        .Ptr(p) => { @assert(p == null, "null ptr payload"); },
+        _ => { @assert(false, "expected ptr"); },
+    }
+}
+
+fn main() {
+    test_int_and_uint();
+    test_scalars();
+    test_borrowed_str();
+    test_ptr();
+}