mirror of
https://github.com/ziglang/zig.git
synced 2025-12-27 16:43:07 +00:00
404 lines
13 KiB
Zig
404 lines
13 KiB
Zig
const builtin = @import("builtin");
|
|
const io = @import("std").io;
|
|
const os = @import("std").os;
|
|
const heap = @import("std").heap;
|
|
const warn = @import("std").debug.warn;
|
|
const assert = @import("std").debug.assert;
|
|
const mem = @import("std").mem;
|
|
const ArrayList = @import("std").ArrayList;
|
|
|
|
|
|
const Token = struct {
|
|
id: Id,
|
|
start: usize,
|
|
end: usize,
|
|
|
|
const KeywordId = struct {
|
|
bytes: []const u8,
|
|
id: Id,
|
|
};
|
|
|
|
const keywords = []KeywordId {
|
|
KeywordId{.bytes="align", .id = Id {.Keyword_align = {}}},
|
|
KeywordId{.bytes="and", .id = Id {.Keyword_and = {}}},
|
|
KeywordId{.bytes="asm", .id = Id {.Keyword_asm = {}}},
|
|
KeywordId{.bytes="break", .id = Id {.Keyword_break = {}}},
|
|
KeywordId{.bytes="coldcc", .id = Id {.Keyword_coldcc = {}}},
|
|
KeywordId{.bytes="comptime", .id = Id {.Keyword_comptime = {}}},
|
|
KeywordId{.bytes="const", .id = Id {.Keyword_const = {}}},
|
|
KeywordId{.bytes="continue", .id = Id {.Keyword_continue = {}}},
|
|
KeywordId{.bytes="defer", .id = Id {.Keyword_defer = {}}},
|
|
KeywordId{.bytes="else", .id = Id {.Keyword_else = {}}},
|
|
KeywordId{.bytes="enum", .id = Id {.Keyword_enum = {}}},
|
|
KeywordId{.bytes="error", .id = Id {.Keyword_error = {}}},
|
|
KeywordId{.bytes="export", .id = Id {.Keyword_export = {}}},
|
|
KeywordId{.bytes="extern", .id = Id {.Keyword_extern = {}}},
|
|
KeywordId{.bytes="false", .id = Id {.Keyword_false = {}}},
|
|
KeywordId{.bytes="fn", .id = Id {.Keyword_fn = {}}},
|
|
KeywordId{.bytes="for", .id = Id {.Keyword_for = {}}},
|
|
KeywordId{.bytes="goto", .id = Id {.Keyword_goto = {}}},
|
|
KeywordId{.bytes="if", .id = Id {.Keyword_if = {}}},
|
|
KeywordId{.bytes="inline", .id = Id {.Keyword_inline = {}}},
|
|
KeywordId{.bytes="nakedcc", .id = Id {.Keyword_nakedcc = {}}},
|
|
KeywordId{.bytes="noalias", .id = Id {.Keyword_noalias = {}}},
|
|
KeywordId{.bytes="null", .id = Id {.Keyword_null = {}}},
|
|
KeywordId{.bytes="or", .id = Id {.Keyword_or = {}}},
|
|
KeywordId{.bytes="packed", .id = Id {.Keyword_packed = {}}},
|
|
KeywordId{.bytes="pub", .id = Id {.Keyword_pub = {}}},
|
|
KeywordId{.bytes="return", .id = Id {.Keyword_return = {}}},
|
|
KeywordId{.bytes="stdcallcc", .id = Id {.Keyword_stdcallcc = {}}},
|
|
KeywordId{.bytes="struct", .id = Id {.Keyword_struct = {}}},
|
|
KeywordId{.bytes="switch", .id = Id {.Keyword_switch = {}}},
|
|
KeywordId{.bytes="test", .id = Id {.Keyword_test = {}}},
|
|
KeywordId{.bytes="this", .id = Id {.Keyword_this = {}}},
|
|
KeywordId{.bytes="true", .id = Id {.Keyword_true = {}}},
|
|
KeywordId{.bytes="undefined", .id = Id {.Keyword_undefined = {}}},
|
|
KeywordId{.bytes="union", .id = Id {.Keyword_union = {}}},
|
|
KeywordId{.bytes="unreachable", .id = Id {.Keyword_unreachable = {}}},
|
|
KeywordId{.bytes="use", .id = Id {.Keyword_use = {}}},
|
|
KeywordId{.bytes="var", .id = Id {.Keyword_var = {}}},
|
|
KeywordId{.bytes="volatile", .id = Id {.Keyword_volatile = {}}},
|
|
KeywordId{.bytes="while", .id = Id {.Keyword_while = {}}},
|
|
};
|
|
|
|
fn getKeyword(bytes: []const u8) -> ?Id {
|
|
for (keywords) |kw| {
|
|
if (mem.eql(u8, kw.bytes, bytes)) {
|
|
return kw.id;
|
|
}
|
|
}
|
|
return null;
|
|
}
|
|
|
|
const StrLitKind = enum {Normal, C};
|
|
|
|
const Id = union(enum) {
|
|
Invalid,
|
|
Identifier,
|
|
StringLiteral: StrLitKind,
|
|
Eof,
|
|
Builtin,
|
|
Equal,
|
|
LParen,
|
|
RParen,
|
|
Semicolon,
|
|
Percent,
|
|
LBrace,
|
|
RBrace,
|
|
Period,
|
|
Minus,
|
|
Arrow,
|
|
Keyword_align,
|
|
Keyword_and,
|
|
Keyword_asm,
|
|
Keyword_break,
|
|
Keyword_coldcc,
|
|
Keyword_comptime,
|
|
Keyword_const,
|
|
Keyword_continue,
|
|
Keyword_defer,
|
|
Keyword_else,
|
|
Keyword_enum,
|
|
Keyword_error,
|
|
Keyword_export,
|
|
Keyword_extern,
|
|
Keyword_false,
|
|
Keyword_fn,
|
|
Keyword_for,
|
|
Keyword_goto,
|
|
Keyword_if,
|
|
Keyword_inline,
|
|
Keyword_nakedcc,
|
|
Keyword_noalias,
|
|
Keyword_null,
|
|
Keyword_or,
|
|
Keyword_packed,
|
|
Keyword_pub,
|
|
Keyword_return,
|
|
Keyword_stdcallcc,
|
|
Keyword_struct,
|
|
Keyword_switch,
|
|
Keyword_test,
|
|
Keyword_this,
|
|
Keyword_true,
|
|
Keyword_undefined,
|
|
Keyword_union,
|
|
Keyword_unreachable,
|
|
Keyword_use,
|
|
Keyword_var,
|
|
Keyword_volatile,
|
|
Keyword_while,
|
|
};
|
|
};
|
|
|
|
const Tokenizer = struct {
|
|
buffer: []const u8,
|
|
index: usize,
|
|
|
|
pub fn dump(self: &Tokenizer, token: &const Token) {
|
|
warn("{} \"{}\"\n", @tagName(token.id), self.buffer[token.start..token.end]);
|
|
}
|
|
|
|
pub fn init(buffer: []const u8) -> Tokenizer {
|
|
return Tokenizer {
|
|
.buffer = buffer,
|
|
.index = 0,
|
|
};
|
|
}
|
|
|
|
const State = enum {
|
|
Start,
|
|
Identifier,
|
|
Builtin,
|
|
C,
|
|
StringLiteral,
|
|
StringLiteralBackslash,
|
|
Minus,
|
|
};
|
|
|
|
pub fn next(self: &Tokenizer) -> Token {
|
|
var state = State.Start;
|
|
var result = Token {
|
|
.id = Token.Id { .Eof = {} },
|
|
.start = self.index,
|
|
.end = undefined,
|
|
};
|
|
while (self.index < self.buffer.len) : (self.index += 1) {
|
|
const c = self.buffer[self.index];
|
|
switch (state) {
|
|
State.Start => switch (c) {
|
|
' ', '\n' => {
|
|
result.start = self.index + 1;
|
|
},
|
|
'c' => {
|
|
state = State.C;
|
|
result.id = Token.Id { .Identifier = {} };
|
|
},
|
|
'"' => {
|
|
state = State.StringLiteral;
|
|
result.id = Token.Id { .StringLiteral = Token.StrLitKind.Normal };
|
|
},
|
|
'a'...'b', 'd'...'z', 'A'...'Z', '_' => {
|
|
state = State.Identifier;
|
|
result.id = Token.Id { .Identifier = {} };
|
|
},
|
|
'@' => {
|
|
state = State.Builtin;
|
|
result.id = Token.Id { .Builtin = {} };
|
|
},
|
|
'=' => {
|
|
result.id = Token.Id { .Equal = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'(' => {
|
|
result.id = Token.Id { .LParen = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
')' => {
|
|
result.id = Token.Id { .RParen = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
';' => {
|
|
result.id = Token.Id { .Semicolon = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'%' => {
|
|
result.id = Token.Id { .Percent = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'{' => {
|
|
result.id = Token.Id { .LBrace = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'}' => {
|
|
result.id = Token.Id { .RBrace = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'.' => {
|
|
result.id = Token.Id { .Period = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'-' => {
|
|
state = State.Minus;
|
|
},
|
|
else => {
|
|
result.id = Token.Id { .Invalid = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
},
|
|
State.Identifier => switch (c) {
|
|
'a'...'z', 'A'...'Z', '_', '0'...'9' => {},
|
|
else => {
|
|
if (Token.getKeyword(self.buffer[result.start..self.index])) |id| {
|
|
result.id = id;
|
|
}
|
|
break;
|
|
},
|
|
},
|
|
State.Builtin => switch (c) {
|
|
'a'...'z', 'A'...'Z', '_', '0'...'9' => {},
|
|
else => break,
|
|
},
|
|
State.C => switch (c) {
|
|
'\\' => @panic("TODO"),
|
|
'"' => {
|
|
state = State.StringLiteral;
|
|
result.id = Token.Id { .StringLiteral = Token.StrLitKind.C };
|
|
},
|
|
'a'...'z', 'A'...'Z', '_', '0'...'9' => {
|
|
state = State.Identifier;
|
|
},
|
|
else => break,
|
|
},
|
|
State.StringLiteral => switch (c) {
|
|
'\\' => {
|
|
state = State.StringLiteralBackslash;
|
|
},
|
|
'"' => {
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
'\n' => break, // Look for this error later.
|
|
else => {},
|
|
},
|
|
|
|
State.StringLiteralBackslash => switch (c) {
|
|
'\n' => break, // Look for this error later.
|
|
else => {
|
|
state = State.StringLiteral;
|
|
},
|
|
},
|
|
|
|
State.Minus => switch (c) {
|
|
'>' => {
|
|
result.id = Token.Id { .Arrow = {} };
|
|
self.index += 1;
|
|
break;
|
|
},
|
|
else => {
|
|
result.id = Token.Id { .Minus = {} };
|
|
break;
|
|
},
|
|
},
|
|
}
|
|
}
|
|
result.end = self.index;
|
|
return result;
|
|
}
|
|
};
|
|
|
|
const AstNode = struct {
|
|
|
|
};
|
|
|
|
const Parser = struct {
|
|
tokenizer: &Tokenizer,
|
|
allocator: &mem.Allocator,
|
|
|
|
fn init(tokenizer: &Tokenizer, allocator: &mem.Allocator) -> Parser {
|
|
return Parser {
|
|
.tokenizer = tokenizer,
|
|
.allocator = allocator,
|
|
};
|
|
}
|
|
|
|
const StackFrame = struct {
|
|
|
|
};
|
|
|
|
const State = enum {
|
|
TopLevel,
|
|
Expression,
|
|
};
|
|
|
|
fn parse(self: &Parser) -> %void {
|
|
var stack = ArrayList(StackFrame).init(self.allocator);
|
|
defer stack.deinit();
|
|
|
|
var state = State.TopLevel;
|
|
while (true) {
|
|
const token = self.tokenizer.next();
|
|
switch (state) {
|
|
State.TopLevel => switch (token.id) {
|
|
Token.Id.Keyword_pub => {
|
|
const next_token = self.tokenizer.next();
|
|
switch (next_token.id) {
|
|
Token.Id.Keyword_fn => {
|
|
const fn_name = self.tokenizer.next();
|
|
if (@TagType(Token.Id)(fn_name.id) != Token.Id.Identifier) {
|
|
@panic("parse error");
|
|
}
|
|
|
|
const lparen = self.tokenizer.next();
|
|
if (@TagType(Token.Id)(lparen.id) != Token.Id.LParen) {
|
|
@panic("parse error");
|
|
}
|
|
},
|
|
Token.Id.Keyword_const => @panic("TODO"),
|
|
Token.Id.Keyword_var => @panic("TODO"),
|
|
Token.Id.Keyword_use => @panic("TODO"),
|
|
else => @panic("parse error"),
|
|
}
|
|
},
|
|
Token.Id.Keyword_const => @panic("TODO"),
|
|
Token.Id.Keyword_var => @panic("TODO"),
|
|
Token.Id.Keyword_fn => @panic("TODO"),
|
|
Token.Id.Keyword_export => @panic("TODO"),
|
|
Token.Id.Keyword_use => @panic("TODO"),
|
|
Token.Id.Keyword_comptime => @panic("TODO"),
|
|
else => @panic("parse error"),
|
|
},
|
|
State.Expression => @panic("TODO"),
|
|
}
|
|
}
|
|
}
|
|
};
|
|
|
|
|
|
pub fn main() -> %void {
|
|
main2() %% |err| {
|
|
warn("{}\n", @errorName(err));
|
|
return err;
|
|
};
|
|
}
|
|
|
|
pub fn main2() -> %void {
|
|
var incrementing_allocator = %return heap.IncrementingAllocator.init(10 * 1024 * 1024);
|
|
defer incrementing_allocator.deinit();
|
|
|
|
const allocator = &incrementing_allocator.allocator;
|
|
|
|
const args = %return os.argsAlloc(allocator);
|
|
defer os.argsFree(allocator, args);
|
|
|
|
const target_file = args[1];
|
|
|
|
const target_file_buf = %return io.readFileAlloc(target_file, allocator);
|
|
|
|
warn("{}", target_file_buf);
|
|
|
|
{
|
|
var tokenizer = Tokenizer.init(target_file_buf);
|
|
while (true) {
|
|
const token = tokenizer.next();
|
|
tokenizer.dump(token);
|
|
if (@TagType(Token.Id)(token.id) == Token.Id.Eof) {
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
|
|
var tokenizer = Tokenizer.init(target_file_buf);
|
|
var parser = Parser.init(&tokenizer, allocator);
|
|
%return parser.parse();
|
|
}
|