codecin 5.4.2__py3-none-any.whl

codecin/native/compiler/syscalls.go

@@ -0,0 +1,86 @@
+// Code generated by script/gen_native_isa.py from codecin/isa.py; DO NOT EDIT.
+// 单一事实来源: codecin/isa.py Syscall 常量.
+
+package compiler
+
+// SYS 功能号 (与 isa.py Syscall 一致; 编译器以 SysXXX 引用)。
+const SysSQRT = 0
+const SysPOW = 1
+const SysABS = 2
+const SysSIN = 3
+const SysCOS = 4
+const SysTAN = 5
+const SysFADD = 6
+const SysFSUB = 7
+const SysFMUL = 8
+const SysFDIV = 9
+const SysFCMP = 10
+const SysFTOI = 11
+const SysITOF = 12
+const SysRAND = 13
+const SysSRAND = 14
+const SysTIME = 15
+const SysSTRLEN = 16
+const SysSTRCMP = 17
+const SysSTRCPY = 18
+const SysSTRCAT = 19
+const SysMALLOC = 20
+const SysPRINT_FLOAT = 21
+const SysITOA = 22
+const SysFTOA = 23
+const SysPRINT_STR = 24
+const SysSTR_CONCAT = 25
+const SysBOOL_STR = 26
+const SysABORT = 27
+const SysSUBSTR = 28
+const SysINDEXOF = 29
+const SysTOUPPER = 30
+const SysTOLOWER = 31
+const SysFLOOR = 32
+const SysCEIL = 33
+const SysROUND = 34
+const SysATOI = 35
+const SysTRIM = 36
+const SysLTRIM = 37
+const SysRTRIM = 38
+const SysAUDIOPLAY = 39
+const SysAUDIOSTOP = 40
+const SysAUDIOVOL = 41
+const SysAUDIOWAIT = 42
+const SysCANVASNEW = 43
+const SysCANVASSET = 44
+const SysCANVASRECT = 45
+const SysCANVASCIRC = 46
+const SysCANVASTEXT = 47
+const SysCANVASLINE = 48
+const SysCANVASSAVE = 49
+const SysCANVASSHOW = 50
+const SysFILEREAD = 51
+const SysFILEWRITE = 52
+const SysFILEAPPEND = 53
+const SysFILEEXISTS = 54
+const SysFILEDELETE = 55
+const SysFILESIZE = 56
+const SysMKDIR = 57
+const SysDIRLIST = 58
+const SysEXEC = 59
+const SysEXECOUTPUT = 60
+const SysGETENV = 61
+const SysSETENV = 62
+const SysOSNAME = 63
+const SysHOSTNAME = 64
+const SysUSERNAME = 65
+const SysCWD = 66
+const SysHOMEDIR = 67
+const SysTERMUXAVAIL = 68
+const SysTERMUXNOTIFY = 69
+const SysTERMUXTOAST = 70
+const SysTERMUXCLIPGET = 71
+const SysTERMUXCLIPSET = 72
+const SysTERMUXBATTERY = 73
+const SysTERMUXVIBRATE = 74
+const SysTERMUXTTS = 75
+const SysTERMUXLOCATION = 76
+const SysTERMUXWIFI = 77
+const SysTERMUXDIALOG = 78
+const SysTERMUXSMS = 79

codecin/native/compiler/tokenizer.go

@@ -0,0 +1,336 @@
+package compiler
+
+import (
+	"fmt"
+	"strconv"
+	"strings"
+)
+
+// ---------------- 词法分析 ----------------
+
+// Token 词法单元。
+type Token struct {
+	kind     string
+	ival     int64
+	fval     float64
+	isFloat  bool
+	sval     string
+	line     int
+	filename string
+}
+
+var keywords = map[string]bool{
+	"struct": true, "function": true, "return": true, "if": true, "else": true,
+	"while": true, "for": true, "do": true, "switch": true, "case": true,
+	"default": true, "break": true, "continue": true, "true": true, "false": true,
+	"int": true, "float": true, "bool": true, "string": true, "void": true,
+	"char": true, "short": true, "long": true, "unsigned": true,
+	"set": true, "add": true, "subtract": true, "multiply": true, "divide": true,
+	"increment": true, "decrement": true,
+}
+
+var baseTypeWords = map[string]bool{
+	"int": true, "float": true, "bool": true, "string": true,
+	"char": true, "short": true, "long": true, "unsigned": true,
+}
+
+var compoundAssign = map[string]string{
+	"PLUSEQ": "+=", "MINUSEQ": "-=", "STAREQ": "*=", "SLASHEQ": "/=",
+	"PERCENTEQ": "%=", "ANDEQ": "&=", "OREQ": "|=", "XOREQ": "^=",
+	"SHLEQ": "<<=", "SHREQ": ">>=",
+}
+
+var singleOps = map[byte]string{
+	'{': "LBRACE", '}': "RBRACE", '(': "LPAREN", ')': "RPAREN",
+	'[': "LBRACKET", ']': "RBRACKET", ',': "COMMA", ';': "SEMI",
+	'+': "PLUS", '-': "MINUS", '*': "STAR", '/': "SLASH", '%': "PERCENT",
+	'=': "ASSIGN", '<': "LT", '>': "GT", '!': "BANG", '.': "DOT",
+	'?': "QUESTION", ':': "COLON", '&': "AMP", '|': "PIPE", '^': "CARET",
+	'~': "TILDE",
+}
+
+var strEscapes = map[byte]byte{
+	'n': '\n', 't': '\t', 'r': '\r', '"': '"', '\\': '\\', '0': 0,
+}
+
+// tokenize 把 CIN 源码切分为 token 流 (含续行处理与 BOM 容忍)。
+func tokenize(source, filename string) ([]Token, error) {
+	source = strings.TrimPrefix(source, "\ufeff") // 容忍 UTF-8 BOM
+	var tokens []Token
+	i, line, n := 0, 1, len(source)
+
+	skipBlock := func() {
+		i += 2
+		for i < n-1 && !(source[i] == '*' && source[i+1] == '/') {
+			if source[i] == '\n' {
+				line++
+			}
+			i++
+		}
+		i += 2
+	}
+
+	for i < n {
+		c := source[i]
+		switch {
+		case c == '\n':
+			tokens = append(tokens, Token{kind: "NL", line: line})
+			line++
+			i++
+		case c == ' ' || c == '\t' || c == '\r':
+			i++
+		case c == '/' && i+1 < n && source[i+1] == '/':
+			for i < n && source[i] != '\n' {
+				i++
+			}
+		case c == '/' && i+1 < n && source[i+1] == '*':
+			skipBlock()
+		case c == '"':
+			startLine := line
+			i++
+			var buf strings.Builder
+			for i < n && source[i] != '"' {
+				if source[i] == '\\' && i+1 < n {
+					esc := source[i+1]
+					if b, ok := strEscapes[esc]; ok {
+						buf.WriteByte(b)
+					} else {
+						buf.WriteByte(esc)
+					}
+					i += 2
+				} else {
+					if source[i] == '\n' {
+						line++
+					}
+					buf.WriteByte(source[i])
+					i++
+				}
+			}
+			i++ // closing quote
+			tokens = append(tokens, Token{kind: "STRING", sval: buf.String(), line: startLine})
+		case c >= '0' && c <= '9' || (c == '.' && i+1 < n && source[i+1] >= '0' && source[i+1] <= '9'):
+			start := i
+			isFloat := false
+			if c == '0' && i+1 < n && strings.IndexByte("xXbBoO", source[i+1]) >= 0 {
+				baseChar := source[i+1]
+				var base int
+				var allowed string
+				switch baseChar {
+				case 'x', 'X':
+					base, allowed = 16, "0123456789abcdefABCDEF_"
+				case 'b', 'B':
+					base, allowed = 2, "01_"
+				default:
+					base, allowed = 8, "01234567_"
+				}
+				i += 2
+				dstart := i
+				for i < n && strings.IndexByte(allowed, source[i]) >= 0 {
+					i++
+				}
+				digits := source[dstart:i]
+				hasDigit := false
+				for _, ch := range digits {
+					if ch != '_' {
+						hasDigit = true
+						break
+					}
+				}
+				if i == dstart || !hasDigit {
+					return nil, fmt.Errorf("Malformed numeric literal at line %d", line)
+				}
+				for i < n && strings.IndexByte("uUlL", source[i]) >= 0 {
+					i++
+				}
+				if i < n && (source[i] == 'f' || source[i] == 'F') {
+					isFloat = true
+					i++
+				}
+				clean := strings.ReplaceAll(digits, "_", "")
+				v, err := strconv.ParseUint(clean, base, 64)
+				if err != nil {
+					return nil, fmt.Errorf("Malformed numeric literal at line %d", line)
+				}
+				_ = start
+				if isFloat {
+					tokens = append(tokens, Token{kind: "FLOAT", fval: float64(v), isFloat: true, line: line})
+				} else {
+					tokens = append(tokens, Token{kind: "NUMBER", ival: int64(v), line: line})
+				}
+			} else {
+				for i < n && (source[i] >= '0' && source[i] <= '9' || source[i] == '.' || source[i] == '_') {
+					if source[i] == '.' {
+						isFloat = true
+					}
+					i++
+				}
+				if i < n && (source[i] == 'e' || source[i] == 'E') {
+					isFloat = true
+					i++
+					if i < n && (source[i] == '+' || source[i] == '-') {
+						i++
+					}
+					for i < n && (source[i] >= '0' && source[i] <= '9' || source[i] == '_') {
+						i++
+					}
+				}
+				digits := source[start:i]
+				for i < n && strings.IndexByte("uUlL", source[i]) >= 0 {
+					i++
+				}
+				if i < n && (source[i] == 'f' || source[i] == 'F') {
+					isFloat = true
+					i++
+				}
+				text := strings.ReplaceAll(digits, "_", "")
+				if isFloat {
+					f, err := strconv.ParseFloat(text, 64)
+					if err != nil {
+						return nil, fmt.Errorf("Malformed numeric literal at line %d", line)
+					}
+					tokens = append(tokens, Token{kind: "FLOAT", fval: f, isFloat: true, line: line})
+				} else {
+					iv, err := strconv.ParseInt(text, 10, 64)
+					if err != nil {
+						return nil, fmt.Errorf("Malformed numeric literal at line %d", line)
+					}
+					tokens = append(tokens, Token{kind: "NUMBER", ival: iv, line: line})
+				}
+			}
+		case c >= 'a' && c <= 'z' || c >= 'A' && c <= 'Z' || c == '_':
+			start := i
+			for i < n && (isAlnum(source[i]) || source[i] == '_') {
+				i++
+			}
+			tokens = append(tokens, Token{kind: "IDENT", sval: source[start:i], line: line})
+		case c == '-' && i+1 < n && source[i+1] == '>':
+			tokens = append(tokens, Token{kind: "ARROW", sval: "->", line: line})
+			i += 2
+		case c == '\'':
+			startLine := line
+			i++
+			if i >= n {
+				return nil, fmt.Errorf("Unterminated char literal at line %d", startLine)
+			}
+			var val int64
+			if source[i] == '\\' && i+1 < n {
+				esc := source[i+1]
+				escMap := map[byte]int64{'n': 10, 't': 9, 'r': 13, '0': 0, 'a': 7, 'b': 8,
+					'f': 12, 'v': 11, '\'': 39, '"': 34, '\\': 92}
+				if v, ok := escMap[esc]; ok {
+					val = v
+				} else {
+					val = int64(esc)
+				}
+				i += 2
+			} else {
+				val = int64(source[i])
+				i++
+			}
+			if i >= n || source[i] != '\'' {
+				return nil, fmt.Errorf("Unterminated char literal at line %d", startLine)
+			}
+			i++
+			tokens = append(tokens, Token{kind: "NUMBER", ival: val, line: line})
+		default:
+			if strings.IndexByte("+-*/%", c) >= 0 && i+1 < n && source[i+1] == '=' {
+				kindMap := map[byte]string{'+': "PLUSEQ", '-': "MINUSEQ", '*': "STAREQ",
+					'/': "SLASHEQ", '%': "PERCENTEQ"}
+				tokens = append(tokens, Token{kind: kindMap[c], sval: string(c) + "=", line: line})
+				i += 2
+			} else if strings.IndexByte("&|^", c) >= 0 && i+1 < n && source[i+1] == '=' {
+				kindMap := map[byte]string{'&': "ANDEQ", '|': "OREQ", '^': "XOREQ"}
+				tokens = append(tokens, Token{kind: kindMap[c], sval: string(c) + "=", line: line})
+				i += 2
+			} else if (c == '<' || c == '>') && i+2 < n && source[i+1] == c && source[i+2] == '=' {
+				k := "SHLEQ"
+				if c == '>' {
+					k = "SHREQ"
+				}
+				tokens = append(tokens, Token{kind: k, sval: string(c) + string(c) + "=", line: line})
+				i += 3
+			} else if (c == '<' || c == '>') && i+1 < n && source[i+1] == c {
+				k := "SHL"
+				if c == '>' {
+					k = "SHR"
+				}
+				tokens = append(tokens, Token{kind: k, sval: string(c) + string(c), line: line})
+				i += 2
+			} else if c == '+' && i+1 < n && source[i+1] == '+' {
+				tokens = append(tokens, Token{kind: "INC", sval: "++", line: line})
+				i += 2
+			} else if c == '-' && i+1 < n && source[i+1] == '-' {
+				tokens = append(tokens, Token{kind: "DEC", sval: "--", line: line})
+				i += 2
+			} else if c == '=' && i+1 < n && source[i+1] == '=' {
+				tokens = append(tokens, Token{kind: "EQ", sval: "==", line: line})
+				i += 2
+			} else if c == '!' && i+1 < n && source[i+1] == '=' {
+				tokens = append(tokens, Token{kind: "NEQ", sval: "!=", line: line})
+				i += 2
+			} else if c == '<' && i+1 < n && source[i+1] == '=' {
+				tokens = append(tokens, Token{kind: "LE", sval: "<=", line: line})
+				i += 2
+			} else if c == '>' && i+1 < n && source[i+1] == '=' {
+				tokens = append(tokens, Token{kind: "GE", sval: ">=", line: line})
+				i += 2
+			} else if c == '&' && i+1 < n && source[i+1] == '&' {
+				tokens = append(tokens, Token{kind: "AND", sval: "&&", line: line})
+				i += 2
+			} else if c == '|' && i+1 < n && source[i+1] == '|' {
+				tokens = append(tokens, Token{kind: "OR", sval: "||", line: line})
+				i += 2
+			} else if k, ok := singleOps[c]; ok {
+				tokens = append(tokens, Token{kind: k, sval: string(c), line: line})
+				i++
+			} else {
+				return nil, fmt.Errorf("Unexpected character %q at line %d", c, line)
+			}
+		}
+	}
+
+	// 续行处理 (圆/方括号内或行尾运算符后的 NL 丢弃)
+	var filtered []Token
+	depth := 0
+	openK := map[string]bool{"LPAREN": true, "LBRACKET": true}
+	closeK := map[string]bool{"RPAREN": true, "RBRACKET": true}
+	continueOps := map[string]bool{
+		"PLUS": true, "MINUS": true, "STAR": true, "SLASH": true, "PERCENT": true,
+		"ASSIGN": true, "LT": true, "GT": true, "LE": true, "GE": true, "EQ": true,
+		"NEQ": true, "AND": true, "OR": true, "COMMA": true, "ARROW": true, "DOT": true,
+		"PLUSEQ": true, "MINUSEQ": true, "STAREQ": true, "SLASHEQ": true,
+		"PERCENTEQ": true, "INC": true, "DEC": true, "AMP": true, "PIPE": true,
+		"CARET": true, "TILDE": true, "SHL": true, "SHR": true, "ANDEQ": true,
+		"OREQ": true, "XOREQ": true, "SHLEQ": true, "SHREQ": true,
+	}
+	for _, tok := range tokens {
+		if openK[tok.kind] {
+			depth++
+		} else if closeK[tok.kind] {
+			if depth > 0 {
+				depth--
+			}
+		}
+		if tok.kind == "NL" {
+			if depth > 0 {
+				continue
+			}
+			if len(filtered) > 0 {
+				prev := filtered[len(filtered)-1]
+				if continueOps[prev.kind] {
+					continue
+				}
+				if prev.kind == "NL" {
+					continue
+				}
+			}
+		}
+		filtered = append(filtered, tok)
+	}
+	filtered = append(filtered, Token{kind: "EOF", line: line})
+	return filtered, nil
+}
+
+func isAlnum(c byte) bool {
+	return c >= '0' && c <= '9' || c >= 'a' && c <= 'z' || c >= 'A' && c <= 'Z'
+}

codecin/native/compiler/types.go

@@ -0,0 +1,213 @@
+// Package compiler 是 Code CIN 语言的 Go 版编译器 (Python codecin/cin.py 的忠实移植)。
+// 它将 CIN 源码编译为 ir.Program, 与 Python CINCompiler 的产物逐字节一致。
+package compiler
+
+import "codecin-native/ir"
+
+// ---------------- 类型系统 ----------------
+
+// TypeKind 类型类别。
+type TypeKind uint8
+
+const (
+	kInt TypeKind = iota
+	kFloat
+	kBool
+	kString
+	kVoid
+	kStruct
+	kArray
+	kPtrArray
+)
+
+// Type 表示一个 CIN 类型。等价于 Python 侧的元组表示:
+//
+//	int/float/bool/string/void  -> Kind 对应标量
+//	('struct', Name)            -> kStruct
+//	('array', elem, size)       -> kArray
+//	('ptrarray', elem)          -> kPtrArray
+type Type struct {
+	Kind TypeKind
+	Elem *Type
+	Size int
+	Name string
+}
+
+func scalarT(k TypeKind) *Type { return &Type{Kind: k} }
+
+func isFixedArray(t *Type) bool { return t != nil && t.Kind == kArray }
+func isPtrArray(t *Type) bool   { return t != nil && t.Kind == kPtrArray }
+func isStruct(t *Type) bool     { return t != nil && t.Kind == kStruct }
+func isScalar(t *Type) bool {
+	return t != nil && (t.Kind == kInt || t.Kind == kFloat || t.Kind == kBool ||
+		t.Kind == kString || t.Kind == kVoid)
+}
+
+func typeSlots(t *Type) int {
+	if isFixedArray(t) {
+		return t.Size * typeSlots(t.Elem)
+	}
+	return 1
+}
+
+func typeName(t *Type) string {
+	if t == nil {
+		return "nil"
+	}
+	switch t.Kind {
+	case kInt:
+		return "int"
+	case kFloat:
+		return "float"
+	case kBool:
+		return "bool"
+	case kString:
+		return "string"
+	case kVoid:
+		return "void"
+	case kStruct:
+		return t.Name
+	case kArray:
+		return typeName(t.Elem) + "[]"
+	case kPtrArray:
+		return typeName(t.Elem) + "[]"
+	}
+	return "?"
+}
+
+// ---------------- AST ----------------
+
+// Node 是 AST 节点 (等价于 Python 侧的位置元组)。
+type Node struct {
+	Kind string // num/bool/str/var/call/binop/not/bitnot/neg/preinc/predec/postinc/postdec/cond/member/index
+	//        block/return/if/while/for/dowhile/switch/break/continue/decl/cpu/assert/expr
+	//        declitem/case/arraylit
+	Num        float64
+	Ival       int64
+	IsFloat    bool
+	Bool       bool
+	Str        string
+	Name       string
+	Op         string
+	A, B, C, D *Node
+	List       []*Node
+	Int        int
+	Type       *Type
+	Filename   string
+	// 数组字面量: Is2D=false 时 ArrayLit[0] 为元素; Is2D=true 时 ArrayLit 为行。
+	ArrayLit [][]*Node
+	Is2D     bool
+	// cpu 语句操作数 (kind, value) 对。
+	Operands [][2]string
+}
+
+// DeclItem 一条变量声明 (name, type, init, arrayLit)。
+type DeclItem struct {
+	Name     string
+	Type     *Type
+	Init     *Node
+	ArrayLit *Node
+}
+
+// SwitchBranch 一个 switch 分支 ('case', constExpr|nil, stmts)。
+type SwitchBranch struct {
+	Const *Node // nil = default
+	Stmts []*Node
+}
+
+// ---------------- 编译上下文 ----------------
+
+// compiler 持有整个编译过程的可变状态。
+type compiler struct {
+	// 词法/语法产物
+	structs   map[string]*StructDef
+	functions map[string]*FuncDef
+	globals   []*GlobalVar
+
+	// 代码生成状态
+	res      *ir.Program
+	dataPtr  int
+	heapStr  map[string]int
+	filename string
+	bounds   bool
+
+	// err 是粘性的首个编译错误 (语义错误不再被静默吞掉, 由 Compile 统一返回)。
+	// 见 docs/SUGGESTIONS.md §3.1.4: 旧实现在语义错误分支只 return/continue,
+	// 会漏发射指令并带着上一次的 x0 继续生成, 产出静默错误的字节码。
+	err error
+
+	// 函数生成上下文
+	funcDef      *FuncDef
+	locals       map[string]localVar
+	frameBytes   int
+	breakLbls    []string
+	continueLbls []string
+	labelCounter int
+	globalsSym   map[string]globalVar
+}
+
+type localVar struct {
+	t       *Type
+	off     int
+	isBlock bool
+}
+
+type globalVar struct {
+	t       *Type
+	addr    int
+	isBlock bool
+}
+
+// StructDef struct 定义。
+type StructDef struct {
+	name      string
+	fields    []fieldDef
+	offsets   map[string]int
+	sizeSlots int
+}
+
+type fieldDef struct {
+	name string
+	t    *Type
+}
+
+// FuncDef 函数定义。
+type FuncDef struct {
+	name    string
+	params  []paramDef
+	retType *Type
+	body    []*Node
+	line    int
+}
+
+type paramDef struct {
+	name string
+	t    *Type
+}
+
+// GlobalVar 全局变量。
+type GlobalVar struct {
+	name     string
+	t        *Type
+	addr     int
+	init     *Node
+	arrayLit *Node
+}
+
+// newCompiler 构建编译器。
+func newCompiler(filename string, bounds bool) *compiler {
+	return &compiler{
+		structs:   map[string]*StructDef{},
+		functions: map[string]*FuncDef{},
+		globals:   []*GlobalVar{},
+		res: &ir.Program{
+			Labels:     map[string]int{},
+			DataLabels: map[string]int{},
+		},
+		heapStr:    map[string]int{},
+		filename:   filename,
+		bounds:     bounds,
+		locals:     map[string]localVar{},
+		globalsSym: map[string]globalVar{},
+	}
+}