Create parse for thp state machine dsl

This commit is contained in:
Araozu 2024-05-20 09:21:51 -05:00
parent c7b429105d
commit 838bd233ac
4 changed files with 194 additions and 6 deletions

View File

@ -120,7 +120,7 @@ const codeHtml = highlightCode(trimAndDedent(code));
</div>
<div>
<div class="p-1 border-b border-white">state</div>
<div class="h-24 p-1">
<div class="h-24 p-1 overflow-y-scroll">
<template x-for="(value, key) in state">
<div x-text="key.replaceAll(' ', '&nbsp;') + ' = ' + value"></div>
</template>
@ -128,10 +128,10 @@ const codeHtml = highlightCode(trimAndDedent(code));
</div>
</div>
<div class="border-t border-white p-1">
<button class="font-mono px-1 rounded bg-pink-950 disabled:opacity-50 disabled:cursor-not-allowed" @click="alpineNext($data)" :disabled="done && 'true'">
<button class="font-mono px-1 rounded bg-pink-200 dark:bg-pink-950 text-black dark:text-white disabled:opacity-50 disabled:cursor-not-allowed" @click="alpineNext($data)" :disabled="done && 'true'">
Step: <span x-text="ip"></span>
</button>
<button class="font-mono px-1 rounded bg-pink-950" @click="alpineReset($data)">
<button class="font-mono px-1 rounded bg-pink-200 dark:bg-pink-950 text-black dark:text-white" @click="alpineReset($data)">
Reset
</button>
</div>

View File

@ -10,11 +10,11 @@ import { is_identifier_char } from "./utils";
* @param is_datatype whether the identifier is a datatype
*/
export function scan_identifier(input: string, starting_position: number, is_datatype = false): [Token, number] {
let value = input[starting_position];
let value = input[starting_position]!;
let pos = starting_position + 1;
while (pos < input.length) {
const c = input[pos];
const c = input[pos]!;
if (is_identifier_char(c)) {
pos += 1;

View File

@ -34,7 +34,7 @@ export function lex(code: string): Array<Token> {
let current_default_token = "";
while (current_pos < code_len) {
const c = code[current_pos];
const c = code[current_pos]!;
// try to scan a number
if (is_digit(c)) {

View File

@ -0,0 +1,188 @@
/*
step {
line 1
set "a" "b"
unset "a"
}
*/
import { scan_number } from "../lexer/number_lexer";
import { scan_string } from "../lexer/string_lexer";
import { is_digit, is_lowercase, is_uppercase } from "../lexer/utils";
enum TokenType {
Step,
Line,
Set,
Number,
String,
Unset,
BraceOpen,
BraceClose,
};
type Token = [TokenType, string | undefined];
// Creates a stream of tokens
function lex(input: string): Array<Token> {
const characters = input.split("");
const characters_len = characters.length;
let next_p = 0;
const tokens: Array<Token> = [];
while (next_p < characters_len)
{
const c = characters[next_p]!;
// word
if (is_lowercase(c) || is_uppercase(c))
{
const [token, next] = lex_word(characters, next_p);
tokens.push(token);
next_p = next;
}
// number
else if (is_digit(c))
{
const [token, next] = scan_number(input, next_p);
tokens.push([TokenType.Number, token.v]);
next_p = next;
}
// string
else if (c === "\"")
{
const [token, next] = scan_string(input, next_p);
tokens.push([TokenType.String, token.v]);
next_p = next;
}
else if (c === "{")
{
tokens.push([TokenType.BraceOpen, undefined]);
next_p += 1;
}
else if (c === "}")
{
tokens.push([TokenType.BraceClose, undefined]);
next_p += 1;
}
else if (c === " " || c === "\n")
{
next_p += 1;
}
else
{
throw new Error(`Invalid character: \`${c}\``);
}
}
return tokens;
}
function lex_word(input: Array<string>, pos: number): [Token, number] {
let next_p = pos;
let value = "";
let c = input[next_p];
while (c !== undefined && (is_lowercase(c) || is_uppercase(c) || is_digit(c) || c === "_"))
{
value += c;
next_p += 1;
c = input[next_p];
}
let token_type;
if (value === "step") { token_type = TokenType.Step; }
else if (value === "line") { token_type = TokenType.Line; }
else if (value === "set") { token_type = TokenType.Set; }
else if (value === "unset"){ token_type = TokenType.Unset; }
else
{
throw new Error(`Invalid word: ${value}`);
}
return [[token_type, value], next_p]
}
enum InstructionType {
Line,
Set,
Unset,
}
type Instruction = {
t: InstructionType,
v0: string,
v1?: string,
}
// Parses the tokens into a instruction set
function parse(tokens: Array<Token>): Array<Array<Instruction>> {
let pos = 0;
let max = tokens.length;
return [];
}
function parse_step(tokens: Array<Token>, _pos: number): Array<Instruction> {
let pos = _pos;
expect(tokens, pos, TokenType.Step, "expected step");
pos += 1;
expect(tokens, pos, TokenType.BraceOpen, "expected opening brace");
pos += 1;
const instructions = [];
while (true) {
const [inst, next] = parse_instruction(tokens, pos);
if (inst === null) {
break;
}
instructions.push(inst);
pos = next;
}
expect(tokens, pos, TokenType.BraceClose, "expected closing brace");
pos += 1
return instructions;
}
function parse_instruction(tokens: Array<Token>, _pos: number): [Instruction|null, number] {
let pos = _pos;
let instruction_type = tokens[pos]![0];
if (instruction_type === TokenType.Line) {
pos += 1;
expect(tokens, pos, TokenType.Number, "expected a number after the `line` instruction");
return [{
t: InstructionType.Line,
v0: tokens[pos]![1]!,
}, pos + 1]
}
else if (instruction_type === TokenType.Set) {
pos += 1;
expect(tokens, pos, TokenType.String, "expected a string after the `set` instruction");
pos += 1;
expect(tokens, pos, TokenType.String, "expected a second string after the `set` instruction");
return [{
t: InstructionType.Set,
v0: tokens[pos - 1]![1]!,
v1: tokens[pos]![1]!,
}, pos + 1]
}
return [null, pos];
}
function expect(t: Array<Token>, pos: number, type: TokenType, err: string) {
const [t_type] = t[pos]!;
if (t_type !== type) {
console.error(t[pos]);
throw new Error(err);
}
}
console.log(parse_step(lex(" step { line 20 set \"a\" \"b\" }"), 0))