import pytest

from core.concept import Concept
from core.sheerka import Sheerka, ExecutionContext
from core.tokenizer import Tokenizer, TokenKind
from parsers.BaseParser import UnexpectedTokenErrorNode
from parsers.BnfParser import BnfParser, UnexpectedEndOfFileError
from parsers.ConceptLexerParser import StrMatch, Optional, ZeroOrMore, OrderedChoice, Sequence, OneOrMore, \
    ConceptLexerParser, ConceptNode, ConceptMatch


def get_context():
    sheerka = Sheerka(skip_builtins_in_db=True)
    sheerka.initialize("mem://")

    return ExecutionContext("sheerka", "xxxx", sheerka)


@pytest.mark.parametrize("expression, expected", [
    ("'str'", StrMatch("str")),
    ("1", StrMatch("1")),
    ("  1", StrMatch("1")),
    (",", StrMatch(",")),
    ("'foo'?", Optional(StrMatch("foo"))),
    ("'foo'*", ZeroOrMore(StrMatch("foo"))),
    ("'foo'+", OneOrMore(StrMatch("foo"))),
    ("1 | 2 | 3", OrderedChoice(StrMatch("1"), StrMatch("2"), StrMatch("3"))),
    ("1|2|3", OrderedChoice(StrMatch("1"), StrMatch("2"), StrMatch("3"))),
    ("1'|' 2 '|' 3", Sequence(StrMatch("1"), StrMatch("|"), StrMatch("2"), StrMatch("|"), StrMatch("3"))),
    ("1 2 'foo'", Sequence(StrMatch("1"), StrMatch("2"), StrMatch("foo"))),
    ("1 2 | 3 4+", OrderedChoice(
        Sequence(StrMatch("1"), StrMatch("2")),
        Sequence(StrMatch("3"), OneOrMore(StrMatch("4"))))),
    ("1 (2 | 3) 4+", Sequence(StrMatch("1"), OrderedChoice(StrMatch("2"), StrMatch("3")), OneOrMore(StrMatch("4")))),
    ("(1|2)+", OneOrMore(OrderedChoice(StrMatch("1"), StrMatch("2")))),
    ("(1 2)+", OneOrMore(Sequence(StrMatch("1"), StrMatch("2")))),
    ("1 *", Sequence(StrMatch("1"), StrMatch("*"))),
    ("1 ?", Sequence(StrMatch("1"), StrMatch("?"))),
    ("1 +", Sequence(StrMatch("1"), StrMatch("+"))),
    ("(1|*) +", Sequence(OrderedChoice(StrMatch("1"), StrMatch("*")), StrMatch("+"))),
    ("1, :&", Sequence(StrMatch("1"), StrMatch(","), StrMatch(":"), StrMatch("&"))),
    ("(1 )", StrMatch("1")),
    ("foo", ConceptMatch("foo")),
    ("foo*", ZeroOrMore(ConceptMatch("foo"))),
    ("foo 'and' bar+", Sequence(ConceptMatch("foo"), StrMatch("and"), OneOrMore(ConceptMatch("bar")))),
    ("foo | bar?", OrderedChoice(ConceptMatch("foo"), Optional(ConceptMatch("bar")))),
    ("'str' = var", Sequence(StrMatch("str"), StrMatch("="), ConceptMatch("var"))),
    ("'str''='var", Sequence(StrMatch("str"), StrMatch("="), ConceptMatch("var"))),
    ("'str'=var", StrMatch("str", rule_name="var")),
    ("'foo'?=var", Optional(StrMatch("foo"), rule_name="var")),
    ("('foo'?)=var", Optional(StrMatch("foo"), rule_name="var")),
    ("'foo'*=var", ZeroOrMore(StrMatch("foo"), rule_name="var")),
    ("('foo'*)=var", ZeroOrMore(StrMatch("foo"), rule_name="var")),
    ("'foo'+=var", OneOrMore(StrMatch("foo"), rule_name="var")),
    ("('foo'+)=var", OneOrMore(StrMatch("foo"), rule_name="var")),
    ("'foo'=var?", Optional(StrMatch("foo", rule_name="var"))),
    ("('foo'=var)?", Optional(StrMatch("foo", rule_name="var"))),
    ("'foo'=var*", ZeroOrMore(StrMatch("foo", rule_name="var"))),
    ("('foo'=var)*", ZeroOrMore(StrMatch("foo", rule_name="var"))),
    ("'foo'=var+", OneOrMore(StrMatch("foo", rule_name="var"))),
    ("('foo'=var)+", OneOrMore(StrMatch("foo", rule_name="var"))),
    ("(1 | 2 | 3)=var", OrderedChoice(StrMatch("1"), StrMatch("2"), StrMatch("3"), rule_name="var")),
    ("(1 2)=var", Sequence(StrMatch("1"), StrMatch("2"), rule_name="var")),
    ("(1 2)+=var", OneOrMore(Sequence(StrMatch("1"), StrMatch("2")), rule_name="var")),
    ("(1 2)=var+", OneOrMore(Sequence(StrMatch("1"), StrMatch("2"), rule_name="var"))),
])
def test_i_can_parse_regex(expression, expected):
    parser = BnfParser()
    res = parser.parse(get_context(), Tokenizer(expression))

    assert not parser.has_error
    assert res.status
    assert res.value.value == expected
    assert res.value.source == expression


@pytest.mark.parametrize("expression, error", [
    ("1 ", UnexpectedEndOfFileError()),
    ("1|", UnexpectedEndOfFileError()),
    ("(1|)", UnexpectedTokenErrorNode("Unexpected token 'Token(<EOF>)'", [TokenKind.RPAR])),
    ("1=", UnexpectedTokenErrorNode("Unexpected token 'Token(<EOF>)'", [TokenKind.IDENTIFIER])),
])
def test_i_can_detect_errors(expression, error):
    parser = BnfParser()
    res = parser.parse(get_context(), Tokenizer(expression))
    ret_value = res.value.value
    assert parser.has_error
    assert not res.status
    assert ret_value[0] == error


def test_i_can_use_the_result_of_regex_parsing_to_parse_a_text():
    foo = Concept(name="foo")
    bar = Concept(name="bar")
    context = get_context()
    context.sheerka.add_in_cache(foo)
    context.sheerka.add_in_cache(bar)

    regex_parser = BnfParser()
    foo_definition = regex_parser.parse(context, "'twenty' | 'thirty'").value.value
    bar_definition = regex_parser.parse(context, "foo ('one' | 'two')").value.value

    concepts = {bar: bar_definition, foo: foo_definition}
    concept_parser = ConceptLexerParser()
    concept_parser.initialize(context, concepts)

    res = concept_parser.parse(context, "twenty two")
    assert res.status
    assert res.value.body == [(bar, 0, 2, "twenty two")]

    res = concept_parser.parse(context, "thirty one")
    assert res.status
    assert res.value.body == [(bar, 0, 2, "thirty one")]

    res = concept_parser.parse(context, "twenty")
    assert res.status
    assert res.value.body == [(foo, 0, 0, "twenty")]