capitalex revidoval tento gist . Přejít na revizi
1 file changed, 133 insertions
alta.lua(vytvořil soubor)
| @@ -0,0 +1,133 @@ | |||
| 1 | + | local function tokenize(string) | |
| 2 | + | local tokens, current_symbol, current_token = {}, {}, {} | |
| 3 | + | ||
| 4 | + | local function push_token() | |
| 5 | + | if #current_token ~= 0 then | |
| 6 | + | table.insert(current_symbol, table.concat(current_token)) | |
| 7 | + | current_token = {} | |
| 8 | + | end | |
| 9 | + | end | |
| 10 | + | ||
| 11 | + | local function push_symbol() | |
| 12 | + | push_token() | |
| 13 | + | if #current_symbol ~= 0 then | |
| 14 | + | table.insert(tokens, {type = "SYMBOL", value = table.concat(current_symbol, " ")}) | |
| 15 | + | current_symbol = {} | |
| 16 | + | end | |
| 17 | + | end | |
| 18 | + | ||
| 19 | + | for char in string:gmatch(".") do | |
| 20 | + | if char == "-" then | |
| 21 | + | push_symbol() | |
| 22 | + | table.insert(tokens, {type = "RULE"}) | |
| 23 | + | elseif char == ">" then | |
| 24 | + | push_symbol() | |
| 25 | + | table.insert(tokens, {type = "ACCEPT", value = "[accept]"}) | |
| 26 | + | elseif char == ":" then | |
| 27 | + | push_symbol() | |
| 28 | + | table.insert(tokens, {type = "SWITCH", value = "[switch]"}) | |
| 29 | + | elseif char == ";" then | |
| 30 | + | push_symbol() | |
| 31 | + | table.insert(tokens, {type = "DONE", value = "[done]"}) | |
| 32 | + | elseif char == "," then | |
| 33 | + | push_symbol() | |
| 34 | + | elseif char == " " or char == "\n" or char == "\t" then | |
| 35 | + | push_token() | |
| 36 | + | else | |
| 37 | + | table.insert(current_token, char) | |
| 38 | + | end | |
| 39 | + | end | |
| 40 | + | ||
| 41 | + | push_symbol() | |
| 42 | + | return tokens | |
| 43 | + | end | |
| 44 | + | ||
| 45 | + | local function here(self) | |
| 46 | + | return self.tokens[self.current] | |
| 47 | + | end | |
| 48 | + | ||
| 49 | + | local function advance(self) | |
| 50 | + | self.current = self.current + 1 | |
| 51 | + | end | |
| 52 | + | ||
| 53 | + | local function read(self) | |
| 54 | + | return self:here(), self:advance() | |
| 55 | + | end | |
| 56 | + | ||
| 57 | + | local function done(self) | |
| 58 | + | return self.current > #self.tokens | |
| 59 | + | end | |
| 60 | + | ||
| 61 | + | local function new_parser(tokens) | |
| 62 | + | return {current = 1, tokens = tokens, here = here, advance = advance, read = read, done = done} | |
| 63 | + | end | |
| 64 | + | ||
| 65 | + | local parse_rules, parse_rule, parse_header | |
| 66 | + | ||
| 67 | + | function parse_header(parser, header, body) | |
| 68 | + | while not parser:done() do | |
| 69 | + | local token = parser:read() | |
| 70 | + | if token.type == "DONE" then | |
| 71 | + | return parse_rules(parser, header, body) | |
| 72 | + | elseif token.type == "RULE" then | |
| 73 | + | return parse_rule(parser, header, body) | |
| 74 | + | elseif token.type == "ACCEPT" then | |
| 75 | + | -- skip this token in this context | |
| 76 | + | else | |
| 77 | + | table.insert(header, token.value) | |
| 78 | + | end | |
| 79 | + | end | |
| 80 | + | return header, rules | |
| 81 | + | end | |
| 82 | + | ||
| 83 | + | function parse_rule(parser, header, body) | |
| 84 | + | -- Starts a new chunk of opcodes | |
| 85 | + | while not parser:done() do | |
| 86 | + | local token = parser:read() | |
| 87 | + | if token.type == "DONE" then | |
| 88 | + | table.insert(body, "[done]") | |
| 89 | + | return parse_rules(parser, header, body) | |
| 90 | + | elseif token.type == "RULE" then | |
| 91 | + | table.insert(body, "[done]") | |
| 92 | + | return parse_rule(parser, header, body) | |
| 93 | + | elseif token.type == "ACCEPT" then | |
| 94 | + | table.insert(body, "[accept]") | |
| 95 | + | elseif token.type == "SYMBOL" and parser:here().type == "SWITCH" then | |
| 96 | + | parser:advance() | |
| 97 | + | table.insert(body, token.value) | |
| 98 | + | table.insert(body, "[switch]") | |
| 99 | + | else | |
| 100 | + | table.insert(body, token.value) | |
| 101 | + | table.insert(body, "[match]") | |
| 102 | + | end | |
| 103 | + | end | |
| 104 | + | table.insert(rules, body) | |
| 105 | + | return header, rules | |
| 106 | + | end | |
| 107 | + | ||
| 108 | + | function parse_rules(parser, header, body) | |
| 109 | + | while not parser:done() do | |
| 110 | + | local token = parser:read() | |
| 111 | + | if token.type == "RULE" and parser:here().type == "ACCEPT" then | |
| 112 | + | parser:advance() | |
| 113 | + | return parse_header(parser, header, body) | |
| 114 | + | elseif token.type == "RULE" then | |
| 115 | + | return parse_rule(parser, header, body) | |
| 116 | + | end | |
| 117 | + | end | |
| 118 | + | return header, rules | |
| 119 | + | end | |
| 120 | + | ||
| 121 | + | local function parse(tokens) | |
| 122 | + | local header, body, bytecode = {}, {}, {} | |
| 123 | + | local parser = new_parser(tokens) | |
| 124 | + | header, rules = parse_rules(parser, header, body) | |
| 125 | + | for _, op in ipairs(header) do | |
| 126 | + | table.insert(bytecode, op) | |
| 127 | + | end | |
| 128 | + | ||
| 129 | + | for _, op in ipairs(body) do | |
| 130 | + | table.insert(bytecode, op) | |
| 131 | + | end | |
| 132 | + | return bytecode | |
| 133 | + | end | |
Novější
Starší