471 lines
18 KiB
Lua
471 lines
18 KiB
Lua
--[[
|
|
This file is part of parser.lua - table based parsing
|
|
Copyright (C) 2019 Soni L.
|
|
|
|
This program is free software: you can redistribute it and/or modify
|
|
it under the terms of the GNU Affero General Public License as published by
|
|
the Free Software Foundation, either version 3 of the License, or
|
|
(at your option) any later version.
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU Affero General Public License for more details.
|
|
|
|
You should have received a copy of the GNU Affero General Public License
|
|
along with this program. If not, see <https://www.gnu.org/licenses/>.
|
|
--]]
|
|
|
|
local parser = require "parser"
|
|
|
|
local caseno = 0
|
|
local function case()
|
|
caseno = caseno + 1
|
|
return caseno
|
|
end
|
|
|
|
do -- basic check
|
|
local case = case()
|
|
local defs = {}
|
|
local count = 0
|
|
local state, err = parser.parse(defs, function() assert(count == 0, "should be called only once"); count = count + 1 return nil end)
|
|
assert(state)
|
|
end -- basic check
|
|
|
|
do -- trim left spaces
|
|
local defs = {}
|
|
defs.self = defs
|
|
defs[' '] = "whitespace"
|
|
defs['\n'] = "whitespace"
|
|
defs['\r'] = "whitespace"
|
|
defs['\t'] = "whitespace"
|
|
defs['\f'] = "whitespace"
|
|
defs['\v'] = "whitespace"
|
|
defs.whitespace = "self"
|
|
defs[''] = function(state, token)
|
|
state[#state + 1] = token
|
|
if #state > 20 then
|
|
state[1] = table.concat(state)
|
|
for i=#state, 2, -1 do
|
|
state[i] = nil
|
|
end
|
|
end
|
|
return "start"
|
|
end
|
|
defs.start = {}
|
|
defs.start.self = defs.start
|
|
defs.start[''] = function(state, token)
|
|
state[#state + 1] = token
|
|
if #state > 20 then
|
|
state[1] = table.concat(state)
|
|
for i=#state, 2, -1 do
|
|
state[i] = nil
|
|
end
|
|
end
|
|
return "self"
|
|
end
|
|
for k,v in ipairs({"hello", " hello", "\t \v \n\r hello"}) do
|
|
local state, err = parser.parse(defs, v)
|
|
local case = case()
|
|
if not state then
|
|
print(case, err)
|
|
else
|
|
assert(table.concat(state) == "hello")
|
|
end
|
|
end
|
|
end -- trim left spaces
|
|
|
|
do -- lua tokens
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [["hello world"]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(state[1] == luatokens.tokens.TK_STRING)
|
|
assert(state[2] == "hello world")
|
|
assert(state[3] == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- lua tokens
|
|
|
|
do -- more lua tokens
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [["\a\b\f\n\r\t\v\\\"\'\z \x41\65\
|
|
"]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(state[1] == luatokens.tokens.TK_STRING)
|
|
assert(state[2] == "\7\8\12\10\13\9\11\92\34\39\65\65\10")
|
|
assert(state[3] == nil)
|
|
assert(state.line == 2)
|
|
end
|
|
end -- lua tokens
|
|
|
|
do -- even more lua tokens
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [["\u{000000000000000000000000000000000000000000000000000000000000041}"]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(state[1] == luatokens.tokens.TK_STRING)
|
|
assert(state[2] == "A")
|
|
assert(state[3] == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- lua tokens
|
|
|
|
do -- even more lua tokens
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [["\u{7F}""\u{80}""\u{7FF}""\u{800}""\u{FFFF}""\u{10000}""\u{1FFFFF}""\u{200000}""\u{3FFFFFF}""\u{4000000}""\u{7FFFFFFF}"]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\127")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\194\128")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\223\191")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\224\160\128")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\239\191\191")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\240\144\128\128")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\247\191\191\191")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\248\136\128\128\128")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\251\191\191\191\191")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\252\132\128\128\128\128")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "\253\191\191\191\191\191")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- lua tokens
|
|
|
|
do -- simple lua tokens
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [[[""]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == "[")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- lua tokens
|
|
|
|
do -- simple long string
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [=[[[]]]=])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- long string
|
|
|
|
do -- long string with depth 1
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [==[[=[]=]]==])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- long string
|
|
|
|
do -- long string with "nested" long string
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [==[[=[[[]]]=]]==])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "[[]]")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 1 or not state.line)
|
|
end
|
|
end -- long string
|
|
|
|
do -- long string edge cases
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, "[==[]=]==][==[]]==][=[] ]=][[\n]][[\n ]]")
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "]=")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "]")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "] ")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == "")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_STRING)
|
|
assert(table.remove(state, 1) == " ")
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 3 or not state.line)
|
|
end
|
|
end -- long string
|
|
|
|
do -- keywords
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [[
|
|
and break do else elseif end
|
|
false for function goto if in
|
|
local nil not or repeat return
|
|
then true until while]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_AND)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_BREAK)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_DO)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_ELSE)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_ELSEIF)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_END)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FALSE)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FOR)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FUNCTION)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_GOTO)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_IF)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_IN)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_LOCAL)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_NIL)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_NOT)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_OR)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_REPEAT)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_RETURN)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_THEN)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_TRUE)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_UNTIL)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_WHILE)
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 4)
|
|
end
|
|
end -- keywords
|
|
|
|
do -- "other tokens"
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [[
|
|
+ - * / % ^ #
|
|
& ~ | << >> //
|
|
== ~= <= >= < > =
|
|
( ) { } [ ] ::
|
|
; : , . .. ...]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == "+")
|
|
assert(table.remove(state, 1) == "-")
|
|
assert(table.remove(state, 1) == "*")
|
|
assert(table.remove(state, 1) == "/")
|
|
assert(table.remove(state, 1) == "%")
|
|
assert(table.remove(state, 1) == "^")
|
|
assert(table.remove(state, 1) == "#")
|
|
assert(table.remove(state, 1) == "&")
|
|
assert(table.remove(state, 1) == "~")
|
|
assert(table.remove(state, 1) == "|")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_SHL)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_SHR)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_IDIV)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_EQ)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_NE)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_LE)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_GE)
|
|
assert(table.remove(state, 1) == "<")
|
|
assert(table.remove(state, 1) == ">")
|
|
assert(table.remove(state, 1) == "=")
|
|
assert(table.remove(state, 1) == "(")
|
|
assert(table.remove(state, 1) == ")")
|
|
assert(table.remove(state, 1) == "{")
|
|
assert(table.remove(state, 1) == "}")
|
|
assert(table.remove(state, 1) == "[")
|
|
assert(table.remove(state, 1) == "]")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_DBCOLON)
|
|
assert(table.remove(state, 1) == ";")
|
|
assert(table.remove(state, 1) == ":")
|
|
assert(table.remove(state, 1) == ",")
|
|
assert(table.remove(state, 1) == ".")
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_CONCAT)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_DOTS)
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 5)
|
|
end
|
|
end -- "other tokens"
|
|
|
|
do -- long comments
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [==[--[[
|
|
|
|
--]]]==])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 3)
|
|
end
|
|
end -- long comments
|
|
|
|
do -- numbers
|
|
local luatokens = require "luatokens"
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, [[
|
|
3 345 0xff 0xBEBADA
|
|
3.0 3.1416 314.16e-2 0.31416E1 34e1
|
|
0x0.1E 0xA23p-4 0X1.921FB54442D18P+1]])
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
else
|
|
-- integers
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_INT)
|
|
assert(table.remove(state, 1) == 3)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_INT)
|
|
assert(table.remove(state, 1) == 345)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_INT)
|
|
assert(table.remove(state, 1) == 0xff)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_INT)
|
|
assert(table.remove(state, 1) == 0xBEBADA)
|
|
-- floats
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 3.0)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 3.1416)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 314.16e-2)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 0.31416E1)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 34e1)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 0.1171875)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == 162.1875)
|
|
assert(table.remove(state, 1) == luatokens.tokens.TK_FLT)
|
|
assert(table.remove(state, 1) == math.pi)
|
|
assert(table.remove(state, 1) == nil)
|
|
assert(state.line == 3)
|
|
end
|
|
end -- numbers
|
|
|
|
do -- FUCK
|
|
local luatokens = require "luatokens"
|
|
local luatokens_file = io.open("./luatokens.lua", "r")
|
|
local tokens = luatokens.defs
|
|
local state, err, etoken, estate = parser.parse(tokens, function() return luatokens_file:read(8192) end)
|
|
local case = case()
|
|
if not state then
|
|
print(case, "---- IN TOKENS ----")
|
|
print(case, err, etoken)
|
|
for i,v in pairs(estate) do
|
|
v = luatokens.reverse_keywords[v] or luatokens.reverse_tokens[v] or v
|
|
print(case, i, v)
|
|
end
|
|
print(case, "---- OUT TOKENS ----")
|
|
end
|
|
end -- FUCK
|