jorts/src/scanner.zig

333 lines
9.1 KiB
Zig

const std = @import("std");
const token = @import("token.zig");
const main = @import("main.zig");
const TokenList = std.ArrayList(token.Token);
const TokenType = token.TokenType;
fn isDigit(char: u8) bool {
return char >= '0' and char <= '9';
}
fn isAlpha(c: u8) bool {
return (c >= 'a' and c <= 'z') or
(c >= 'A' and c <= 'Z') or
c == '_';
}
fn isAlphaNumeric(char: u8) bool {
return isAlpha(char) or isDigit(char);
}
// hashmaps don't work on HashMaps for some reason. anyways.
pub const KeywordMap = std.AutoHashMap([]const u8, u6);
fn initKeywordMap(allocator: *std.mem.Allocator) !KeywordMap {
var map = KeywordMap.init(allocator);
const keywords = [][]const u8{
"and"[0..],
"class"[0..],
"else"[0..],
"false"[0..],
"for"[0..],
"fun"[0..],
"if"[0..],
"nil"[0..],
"or"[0..],
"print"[0..],
"return"[0..],
"super"[0..],
"this"[0..],
"true"[0..],
"var"[0..],
"while"[0..],
};
const tags = []TokenType{
TokenType.AND,
TokenType.CLASS,
TokenType.ELSE,
TokenType.FALSE,
TokenType.FOR,
TokenType.FUN,
TokenType.IF,
TokenType.NIL,
TokenType.OR,
TokenType.PRINT,
TokenType.RETURN,
TokenType.SUPER,
TokenType.THIS,
TokenType.TRUE,
TokenType.VAR,
TokenType.WHILE,
};
for (keywords) |keyword, idx| {
var tag = @enumToInt(tags[idx]);
_ = try map.put(keyword, tag);
}
return map;
}
pub const Scanner = struct {
source: []u8,
tokens: TokenList,
keywords: KeywordMap,
start: usize = 0,
current: usize = 0,
line: usize = 1,
pub fn init(allocator: *std.mem.Allocator, data: []u8) !Scanner {
return Scanner{
.source = data,
.tokens = TokenList.init(allocator),
.keywords = try initKeywordMap(allocator),
};
}
fn isAtEnd(self: *Scanner) bool {
return self.current >= self.source.len;
}
fn advance(self: *Scanner) u8 {
self.current += 1;
return self.source[self.current - 1];
}
fn currentLexeme(self: *Scanner) []u8 {
return self.source[self.start..self.current];
}
fn addSimpleToken(self: *Scanner, ttype: token.TokenType) !void {
try self.addToken(token.Token{
.Simple = token.SimpleToken.init(
ttype,
self.currentLexeme(),
self.line,
{},
),
});
}
fn addSliceToken(self: *Scanner, ttype: token.TokenType, slice: []u8) !void {
try self.addToken(token.Token{
.Slice = token.SliceToken.init(
ttype,
self.currentLexeme(),
self.line,
slice,
),
});
}
/// Keep in mind Lox only has a single number type and that is a float one.
fn addNumberToken(self: *Scanner, ttype: token.TokenType, num: f32) !void {
try self.addToken(token.Token{
.Number = token.NumberToken.init(
ttype,
self.currentLexeme(),
self.line,
num,
),
});
}
fn addToken(
self: *Scanner,
tok: token.Token,
) !void {
try self.tokens.append(tok);
}
/// Check if the next character matches what is expected.
fn match(self: *Scanner, expected: u8) bool {
if (self.isAtEnd()) return false;
if (self.source[self.current] != expected) return false;
self.current += 1;
return true;
}
/// Add a SimpleToken of type_match if the next character is
/// `expected`. Adds a SimpleToken of type_nomatch when it is not.
fn addMatchToken(
self: *Scanner,
expected: u8,
type_match: token.TokenType,
type_nomatch: token.TokenType,
) !void {
if (self.match(expected)) {
try self.addSimpleToken(type_match);
} else {
try self.addSimpleToken(type_nomatch);
}
}
fn peek(self: *Scanner) u8 {
if (self.isAtEnd()) return 0;
return self.source[self.current];
}
fn doString(self: *Scanner) !void {
// consume entire string
while (self.peek() != '"' and !self.isAtEnd()) {
if (self.peek() == '\n') self.line += 1;
_ = self.advance();
}
// unterminated string.
if (self.isAtEnd()) {
try main.doError(self.line, "Unterminated string.");
return;
}
// the closing ".
_ = self.advance();
// trim the surrounding quotes.
try self.addSliceToken(
.STRING,
self.source[self.start + 1 .. self.current - 1],
);
}
fn peekNext(self: *Scanner) u8 {
if (self.current + 1 >= self.source.len) return 0;
return self.source[self.current + 1];
}
/// Consume a number
fn doNumber(self: *Scanner) !void {
while (isDigit(self.peek())) {
_ = self.advance();
}
// check if its a number like 12.34, where the '.' character
// exists and the one next to it is a digit.
if (self.peek() == '.' and isDigit(self.peekNext())) {
_ = self.advance();
while (isDigit(self.peek())) {
_ = self.advance();
}
}
// after going through all of the number, we can just use fmt.parseFloat
var num = try std.fmt.parseFloat(
f32,
self.source[self.start..self.current],
);
try self.addNumberToken(.NUMBER, num);
}
fn doIdentifier(self: *Scanner) !void {
while (isAlphaNumeric(self.peek())) {
_ = self.advance();
}
// after reading the identifier, we check
// if it is any of our keywords, if it is, then we add
// the specificed keyword type. if not, just .IDENTIFIER
var text = self.source[self.start..self.current];
var type_opt = self.keywords.get(text);
var toktype: TokenType = undefined;
if (type_opt) |kv| {
toktype = @intToEnum(TokenType, kv.value);
} else {
toktype = TokenType.IDENTIFIER;
}
try self.addSimpleToken(toktype);
}
/// Scan through our tokens and add them to the Scanner's token list.
fn scanToken(self: *Scanner) !void {
var c = self.advance();
switch (c) {
'(' => try self.addSimpleToken(.LEFT_PAREN),
')' => try self.addSimpleToken(.RIGHT_PAREN),
'{' => try self.addSimpleToken(.LEFT_BRACE),
'}' => try self.addSimpleToken(.RIGHT_BRACE),
',' => try self.addSimpleToken(.COMMA),
'.' => try self.addSimpleToken(.DOT),
'-' => try self.addSimpleToken(.MINUS),
'+' => try self.addSimpleToken(.PLUS),
';' => try self.addSimpleToken(.SEMICOLON),
'*' => try self.addSimpleToken(.STAR),
'!' => try self.addMatchToken('=', .BANG_EQUAL, .BANG),
'=' => try self.addMatchToken('=', .EQUAL_EQUAL, .EQUAL),
'<' => try self.addMatchToken('=', .LESS_EQUAL, .LESS),
'>' => try self.addMatchToken('=', .GREATER_EQUAL, .GREATER),
'/' => blk: {
// consume comments
if (self.match('/')) {
while (self.peek() != '\n' and !self.isAtEnd()) {
_ = self.advance();
}
} else if (self.match('*')) {
// multiline block comments are messier to work with, but
// we can still do it!
while (true) {
if (self.isAtEnd()) break;
// check '*/'
if (self.peek() == '*' and self.peekNext() == '/') {
self.current += 2;
break;
}
_ = self.advance();
}
} else {
try self.addSimpleToken(.SLASH);
}
},
' ', '\r', '\t' => blk: {},
'\n' => blk: {
self.line += 1;
},
'"' => try self.doString(),
else => {
if (isDigit(c)) {
try self.doNumber();
} else if (isAlpha(c)) {
try self.doIdentifier();
} else {
try main.doError(self.line, "Unexpected character");
}
},
}
}
pub fn scanTokens(self: *Scanner) !TokenList {
// while we aren't at the end, we're still consuming
// tokens.
while (!self.isAtEnd()) {
self.start = self.current;
try self.scanToken();
}
try self.addToken(token.Token{
.Simple = token.SimpleToken.init(
.EOF,
"",
self.line,
{},
),
});
return self.tokens;
}
};