First steps of ConceptLexer. Need to update DefaultParser before continuing
This commit is contained in:
@@ -8,7 +8,7 @@ from parsers.BaseParser import BaseParser
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
@@ -0,0 +1,560 @@
|
||||
import pytest
|
||||
from core.builtin_concepts import BuiltinConcepts
|
||||
from core.concept import Concept
|
||||
from core.sheerka import Sheerka, ExecutionContext
|
||||
from parsers.ConceptLexerParser import ConceptLexerParser, ConceptNode, Sequence, StrMatch, OrderedChoice, Optional, \
|
||||
CrossRef
|
||||
|
||||
|
||||
@pytest.mark.parametrize("match, text", [
|
||||
("foo", "foo"),
|
||||
("'foo'", "'foo'"),
|
||||
("1", "1"),
|
||||
("3.14", "3.14"),
|
||||
("+", "+"),
|
||||
(StrMatch("foo"), "foo"),
|
||||
(StrMatch("'foo'"), "'foo'"),
|
||||
(StrMatch("1"), "1"),
|
||||
(StrMatch("3.14"), "3.14"),
|
||||
(StrMatch("+"), "+"),
|
||||
])
|
||||
def test_i_can_match_simple_tokens(match, text):
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: text}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, text)
|
||||
|
||||
assert res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.value == [ConceptNode(foo, 0, 0, source=text)]
|
||||
|
||||
|
||||
def test_i_can_match_multiple_concepts_in_one_input():
|
||||
context = get_context()
|
||||
one = Concept(name="one")
|
||||
two = Concept(name="two")
|
||||
concepts = {one: "one", two: "two"}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two one")
|
||||
|
||||
assert res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.value == [
|
||||
ConceptNode(one, 0, 0, source="one"),
|
||||
ConceptNode(two, 2, 2, source="two"),
|
||||
ConceptNode(one, 4, 4, source="one"),
|
||||
]
|
||||
|
||||
|
||||
def test_i_cannot_match_an_unknown_input():
|
||||
context = get_context()
|
||||
parser = ConceptLexerParser() # no grammar registered
|
||||
|
||||
res = parser.parse(context, "foo")
|
||||
|
||||
assert not res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert context.sheerka.isinstance(res.value.body[0], BuiltinConcepts.UNKNOWN_CONCEPT)
|
||||
assert res.value.body[0].body == "foo"
|
||||
|
||||
|
||||
def test_i_cannot_match_when_part_of_the_input_is_unknown():
|
||||
context = get_context()
|
||||
one = Concept(name="one")
|
||||
two = Concept(name="two")
|
||||
concepts = {one: "one", two: "two"}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three")
|
||||
assert not res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.try_parsed == [
|
||||
ConceptNode(one, 0, 0, source="one"),
|
||||
ConceptNode(two, 2, 2, source="two")] # these two were recognized
|
||||
assert context.sheerka.isinstance(res.value.body[0], BuiltinConcepts.UNKNOWN_CONCEPT)
|
||||
assert res.value.body[0].body == "three"
|
||||
|
||||
|
||||
def test_i_can_match_sequence():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: Sequence("one", "two", "three")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three")
|
||||
|
||||
assert res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.value == [ConceptNode(foo, 0, 4, source="one two three")]
|
||||
|
||||
|
||||
def test_wrong_sequence_is_not_matched():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: Sequence("one", "two", "three")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three one")
|
||||
|
||||
assert not res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.try_parsed == [ConceptNode(foo, 0, 4, source="one two three")]
|
||||
assert context.sheerka.isinstance(res.value.body[0], BuiltinConcepts.UNKNOWN_CONCEPT)
|
||||
assert res.value.body[0].body == "one"
|
||||
|
||||
|
||||
def test_i_cannot_match_sequence_if_end_of_file():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: Sequence("one", "two", "three")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two")
|
||||
assert not res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.try_parsed == []
|
||||
assert context.sheerka.isinstance(res.value.body[0], BuiltinConcepts.UNKNOWN_CONCEPT)
|
||||
assert res.value.body[0].body == "one"
|
||||
|
||||
|
||||
def test_i_always_choose_the_longest_match():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
concepts = {bar: Sequence("one", "two"), foo: Sequence("one", "two", "three")}
|
||||
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three")
|
||||
|
||||
assert res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.value == [ConceptNode(foo, 0, 4, source="one two three")]
|
||||
|
||||
|
||||
def test_i_can_match_several_sequences():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
concepts = {bar: Sequence("one", "two"), foo: Sequence("one", "two", "three")}
|
||||
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three one two")
|
||||
|
||||
assert res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res.value.value == [
|
||||
ConceptNode(foo, 0, 4, source="one two three"),
|
||||
ConceptNode(bar, 6, 8, source="one two"),
|
||||
]
|
||||
|
||||
|
||||
def test_i_can_match_ordered_choice():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: OrderedChoice("one", "two")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res1 = parser.parse(context, "one")
|
||||
assert res1.status
|
||||
assert context.sheerka.isinstance(res1.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res1.value.body == [ConceptNode(foo, 0, 0, source="one")]
|
||||
|
||||
res2 = parser.parse(context, "two")
|
||||
assert res2.status
|
||||
assert context.sheerka.isinstance(res2.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res2.value.body == [ConceptNode(foo, 0, 0, source="two")]
|
||||
|
||||
res3 = parser.parse(context, "three")
|
||||
assert not res3.status
|
||||
assert context.sheerka.isinstance(res3.value.body[0], BuiltinConcepts.UNKNOWN_CONCEPT)
|
||||
assert res3.value.body[0].body == "three"
|
||||
|
||||
|
||||
def test_i_cannot_match_ordered_choice_with_empty_alternative():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
concepts = {foo: Sequence(OrderedChoice("one", ""), "two")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "ok") # because token[0] is not "one" and not "" (it is 'two')
|
||||
assert not res.status
|
||||
|
||||
|
||||
def test_i_can_mix_sequences_and_ordered_choices():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Sequence(OrderedChoice("twenty", "thirty"), "one", "ok")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res1 = parser.parse(context, "twenty one ok")
|
||||
assert res1.status
|
||||
assert context.sheerka.isinstance(res1.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res1.value.body == [ConceptNode(foo, 0, 4, source="twenty one ok")]
|
||||
|
||||
res2 = parser.parse(context, "thirty one ok")
|
||||
assert res2.status
|
||||
assert context.sheerka.isinstance(res2.value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res2.value.body == [ConceptNode(foo, 0, 4, source="thirty one ok")]
|
||||
|
||||
res3 = parser.parse(context, "twenty one")
|
||||
assert not res3.status
|
||||
assert res3.value.body[0].body == "twenty"
|
||||
assert res3.value.try_parsed == []
|
||||
|
||||
|
||||
def test_i_can_mix_ordered_choices_and_sequences():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: OrderedChoice(Sequence("twenty", "thirty"), "one")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "twenty thirty")
|
||||
assert res.status
|
||||
|
||||
res = parser.parse(context, "one")
|
||||
assert res.status
|
||||
|
||||
|
||||
def test_i_cannot_parse_empty_optional():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Optional("one")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "")
|
||||
assert not res.status
|
||||
assert context.sheerka.isinstance(res.value, BuiltinConcepts.IS_EMPTY)
|
||||
|
||||
|
||||
def test_i_can_parse_optional():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Optional("one")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one")
|
||||
assert res.status
|
||||
assert res.value.value == [ConceptNode(foo, 0, 0, source="one")]
|
||||
|
||||
|
||||
def test_i_can_parse_sequence_starting_with_optional():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Sequence(Optional("twenty"), "one")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "twenty one")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 2, source="twenty one")]
|
||||
|
||||
res = parser.parse(context, "one")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 0, source="one")]
|
||||
|
||||
|
||||
def test_i_can_parse_sequence_ending_with_optional():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Sequence("one", "two", Optional("three"))}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 4, source="one two three")]
|
||||
|
||||
res = parser.parse(context, "one two")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 2, source="one two")]
|
||||
|
||||
|
||||
def test_i_can_parse_sequence_with_optional_in_between():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
|
||||
concepts = {foo: Sequence("one", Optional("two"), "three")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two three")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 4, source="one two three")]
|
||||
|
||||
res = parser.parse(context, "one three")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 2, source="one three")]
|
||||
|
||||
|
||||
def test_i_can_use_reference():
|
||||
# The problem here is when there are multiple match for the same input
|
||||
# The parsing result is a list of all concepts found
|
||||
# So it's already a list that represents a sequence, not a choice
|
||||
# So I need to create a choice concept
|
||||
# create the return value for every possible graph
|
||||
# --> The latter seems to be the best as we don't defer the resolution of the problem to someone else
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {foo: Sequence("one", "two"), bar: foo}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two")
|
||||
assert len(res) == 2
|
||||
|
||||
assert res[0].status
|
||||
assert context.sheerka.isinstance(res[0].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[0].value.body == [ConceptNode(foo, 0, 2, source="one two")]
|
||||
|
||||
assert res[1].status
|
||||
assert context.sheerka.isinstance(res[1].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[1].value.body == [ConceptNode(bar, 0, 2, source="one two")]
|
||||
|
||||
|
||||
def test_i_can_use_context_reference_with_multiple_levels():
|
||||
"""
|
||||
Same than previous one, but with reference of reference
|
||||
:return:
|
||||
"""
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
baz = Concept(name="baz")
|
||||
|
||||
concepts = {foo: Sequence("one", "two"), bar: foo, baz: bar}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two")
|
||||
assert len(res) == 3
|
||||
|
||||
assert res[0].status
|
||||
assert context.sheerka.isinstance(res[0].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[0].value.body == [ConceptNode(foo, 0, 2, source="one two")]
|
||||
|
||||
assert res[1].status
|
||||
assert context.sheerka.isinstance(res[1].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[1].value.body == [ConceptNode(bar, 0, 2, source="one two")]
|
||||
|
||||
assert res[2].status
|
||||
assert context.sheerka.isinstance(res[2].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[2].value.body == [ConceptNode(baz, 0, 2, source="one two")]
|
||||
|
||||
|
||||
def test_order_is_not_important_when_using_references():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {bar: foo, foo: Sequence("one", "two")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "one two")
|
||||
assert len(res) == 2
|
||||
assert res[0].value.body == [ConceptNode(bar, 0, 2, source="one two")]
|
||||
assert res[1].value.body == [ConceptNode(foo, 0, 2, source="one two")]
|
||||
|
||||
|
||||
def test_i_can_parse_when_reference():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {bar: Sequence(foo, OrderedChoice("one", "two")), foo: OrderedChoice("twenty", "thirty")}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "twenty two")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(bar, 0, 2, source="twenty two")]
|
||||
|
||||
res = parser.parse(context, "thirty one")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(bar, 0, 2, source="thirty one")]
|
||||
|
||||
res = parser.parse(context, "twenty")
|
||||
assert res.status
|
||||
assert res.value.body == [ConceptNode(foo, 0, 0, source="twenty")]
|
||||
|
||||
|
||||
def test_i_can_detect_duplicates_when_reference():
|
||||
context = get_context()
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {
|
||||
bar: Sequence(foo, Optional(OrderedChoice("one", "two"))),
|
||||
foo: OrderedChoice("twenty", "thirty")
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
res = parser.parse(context, "twenty")
|
||||
assert len(res) == 2
|
||||
assert res[0].status
|
||||
assert context.sheerka.isinstance(res[0].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[0].value.body == [ConceptNode(bar, 0, 0, source="twenty")]
|
||||
|
||||
assert res[1].status
|
||||
assert context.sheerka.isinstance(res[1].value, BuiltinConcepts.PARSER_RESULT)
|
||||
assert res[1].value.body == [ConceptNode(foo, 0, 0, source="twenty")]
|
||||
|
||||
|
||||
def test_i_can_detect_infinite_recursion():
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {
|
||||
bar: foo,
|
||||
foo: bar
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
assert bar not in parser.concepts_dict
|
||||
assert foo not in parser.concepts_dict
|
||||
|
||||
|
||||
def test_i_can_detect_indirect_infinite_recursion_with_ordered_choice():
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {
|
||||
bar: foo,
|
||||
foo: OrderedChoice(bar, "foo")
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
assert foo not in parser.concepts_dict # removed because of the infinite recursion
|
||||
assert bar not in parser.concepts_dict # removed because of the infinite recursion
|
||||
|
||||
# the other way around is possible
|
||||
context = get_context()
|
||||
concepts = {
|
||||
bar: foo,
|
||||
foo: OrderedChoice("foo", bar)
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
assert foo in parser.concepts_dict
|
||||
assert bar in parser.concepts_dict
|
||||
|
||||
res = parser.parse(context, "foo")
|
||||
assert len(res) == 2
|
||||
assert res[0].status
|
||||
assert res[0].value.body == [ConceptNode(bar, 0, 0, source="foo")]
|
||||
assert res[1].status
|
||||
assert res[1].value.body == [ConceptNode(foo, 0, 0, source="foo")]
|
||||
|
||||
|
||||
def test_i_can_detect_indirect_infinite_recursion_with_sequence():
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {
|
||||
bar: foo,
|
||||
foo: Sequence("one", bar, "two")
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
assert foo not in parser.concepts_dict # removed because of the infinite recursion
|
||||
assert bar not in parser.concepts_dict # removed because of the infinite recursion
|
||||
|
||||
|
||||
def test_i_can_detect_indirect_infinite_recursion_with_sequence_or_ordered_choice():
|
||||
foo = Concept(name="foo")
|
||||
bar = Concept(name="bar")
|
||||
|
||||
concepts = {
|
||||
bar: foo,
|
||||
foo: Sequence("one", OrderedChoice(bar, "other"), "two")
|
||||
}
|
||||
parser = ConceptLexerParser()
|
||||
parser.initialize(concepts)
|
||||
|
||||
assert foo not in parser.concepts_dict # removed because of the infinite recursion
|
||||
assert bar not in parser.concepts_dict # removed because of the infinite recursion
|
||||
|
||||
|
||||
def test_i_can_detect_indirect_infinite_recursion_with_optional():
|
||||
# TODO infinite recursion with optional
|
||||
pass
|
||||
|
||||
#
|
||||
# def test_i_can_parse_basic_arithmetic_operations_and_resolve_properties():
|
||||
# context = get_context()
|
||||
# add = Concept(name="add")
|
||||
# mult = Concept(name="mult")
|
||||
# atom = Concept(name="atom")
|
||||
#
|
||||
# concepts = {
|
||||
# add: Sequence(mult, Optional(Sequence(OrderedChoice('+', '-', rule_name="sign"), add))),
|
||||
# mult: Sequence(atom, Optional(Sequence(OrderedChoice('*', '/'), mult))),
|
||||
# atom: OrderedChoice(OrderedChoice('1', '2', '3'), Sequence('(', add, ')')),
|
||||
# }
|
||||
#
|
||||
# parser = ConceptLexerParser()
|
||||
# parser.register(concepts)
|
||||
#
|
||||
# # res = parser.parse(context, "1")
|
||||
# # assert len(res) == 3 # add, mult, atom
|
||||
# #
|
||||
# # res = parser.parse(context, "1 * 2")
|
||||
# # assert len(res) == 2 # add and mult
|
||||
# #
|
||||
# # res = parser.parse(context, "1 + 2")
|
||||
# # assert res.status
|
||||
# # assert res.value.value == [ConceptNode(add, 0, 4, source="1 + 2")]
|
||||
#
|
||||
# res = parser.parse(context, "1 * 2 + 3")
|
||||
# assert res.status
|
||||
# assert res.value.value == [ConceptNode(add, 0, 4, source="1 + 2 + 3")]
|
||||
|
||||
|
||||
def test_i_can_register_concepts_with_the_same_name():
|
||||
# TODO : concepts are registered by name,
|
||||
# what when two concepts have the same name ?
|
||||
pass
|
||||
|
||||
|
||||
def test_i_can_parse_very_very_long_input():
|
||||
# TODO: In the current implementation, all the tokens are loaded in memory
|
||||
# It's clearly not the good approach
|
||||
pass
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
|
||||
return ExecutionContext("sheerka", "xxxx", sheerka)
|
||||
@@ -67,7 +67,7 @@ def get_concept(name, where=None, pre=None, post=None, body=None):
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
@@ -125,7 +125,7 @@ def test_i_can_detect_concept_from_tokens():
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
|
||||
return ExecutionContext("sheerka", "xxxx", sheerka)
|
||||
|
||||
@@ -7,7 +7,7 @@ from parsers.BaseParser import BaseParser
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
@@ -8,7 +8,7 @@ from parsers.PythonParser import PythonNode, PythonParser
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
@@ -9,7 +9,7 @@ from parsers.PythonParser import PythonNode, PythonParser, PythonErrorNode
|
||||
|
||||
|
||||
def get_context():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
+29
-4
@@ -8,7 +8,7 @@ from core.sheerka import Sheerka
|
||||
|
||||
|
||||
def get_sheerka():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
|
||||
return sheerka
|
||||
@@ -34,7 +34,7 @@ def my_function(a,b):
|
||||
return a
|
||||
"""
|
||||
tree = ast.parse(source)
|
||||
tree_as_concept = core.ast.nodes.transform(tree)
|
||||
tree_as_concept = core.ast.nodes.python_to_concept(tree)
|
||||
sheerka = get_sheerka()
|
||||
|
||||
assert tree_as_concept.node_type == "Module"
|
||||
@@ -87,7 +87,7 @@ def my_function(a,b):
|
||||
"""
|
||||
|
||||
node = ast.parse(source)
|
||||
concept_node = core.ast.nodes.transform(node)
|
||||
concept_node = core.ast.nodes.python_to_concept(node)
|
||||
|
||||
visitor = TestNameVisitor()
|
||||
visitor.visit(concept_node)
|
||||
@@ -115,7 +115,7 @@ my_function(x,y)
|
||||
sheerka = get_sheerka()
|
||||
|
||||
node = ast.parse(source)
|
||||
concept_node = core.ast.nodes.transform(node)
|
||||
concept_node = core.ast.nodes.python_to_concept(node)
|
||||
|
||||
visitor = UnreferencedNamesVisitor(sheerka)
|
||||
visitor.visit(concept_node)
|
||||
@@ -129,3 +129,28 @@ my_function(x,y)
|
||||
def test_i_can_compare_NodeParent_with_tuple():
|
||||
node_parent = NodeParent(GenericNodeConcept("For", None), "target")
|
||||
assert node_parent == ("For", "target")
|
||||
|
||||
|
||||
def test_i_can_transform_back():
|
||||
source = """
|
||||
def my_function(a,b):
|
||||
for i in range(b):
|
||||
a = a + b
|
||||
return a
|
||||
|
||||
|
||||
my_function(x, y)
|
||||
"""
|
||||
|
||||
node = ast.parse(source)
|
||||
concept_node = core.ast.nodes.python_to_concept(node)
|
||||
|
||||
transformed_back = core.ast.nodes.concept_to_python(concept_node)
|
||||
assert dump_ast(transformed_back) == dump_ast(node)
|
||||
|
||||
|
||||
def dump_ast(node):
|
||||
dump = ast.dump(node)
|
||||
for to_remove in [", ctx=Load()", ", kind=None", ", type_ignores=[]"]:
|
||||
dump = dump.replace(to_remove, "")
|
||||
return dump
|
||||
|
||||
@@ -1,3 +1,7 @@
|
||||
import ast
|
||||
|
||||
import pytest
|
||||
|
||||
from core.builtin_concepts import ReturnValueConcept, BuiltinConcepts
|
||||
from core.sheerka import Sheerka, ExecutionContext
|
||||
import core.builtin_helpers
|
||||
@@ -110,8 +114,47 @@ def test_i_can_use_expect_one_when_not_a_list_false():
|
||||
assert res == item
|
||||
|
||||
|
||||
@pytest.mark.parametrize("expression, vars_to_include, vars_to_exclude, expected_expr", [
|
||||
("a == 1", [], [], []),
|
||||
("a == 1", ["a"], [], ["a == 1"]),
|
||||
("a == 1", [], ["a"], []),
|
||||
("predicate(a)", [], [], []),
|
||||
("predicate(a)", ["a"], [], ["predicate(a)"]),
|
||||
("predicate(a, b)", ["a"], [], ["predicate(a, b)"]),
|
||||
("predicate(a, b)", ["b"], [], ["predicate(a, b)"]),
|
||||
("predicate(a, b)", ["a", "b"], [], ["predicate(a, b)"]),
|
||||
("predicate(a, b)", ["a"], ["b"], []),
|
||||
("a + b == 1", [], [], []),
|
||||
("a + b == 1", ["a"], [], ["a + b == 1"]),
|
||||
("a + b == 1", ["a"], ["b"], []),
|
||||
("a + b == 1", ["b"], [], ["a + b == 1"]),
|
||||
("a + b == 1", ["a", "b"], [], ["a + b == 1"]),
|
||||
("a == 1 and b == 2", [], [], []),
|
||||
("a == 1 and b == 2", ["a"], [], ["a == 1"]),
|
||||
("a == 1 and b == 2", ["b"], [], ["b == 2"]),
|
||||
("a == 1 and b == 2", ["a"], ["b"], ["a == 1"]),
|
||||
("a == 1 and b == 2", ["a", "b"], [], ["a == 1 and b == 2"]),
|
||||
("predicate(a,c) and predicate(b,c)", ["a", "b"], [], ["predicate(a,c) and predicate(b,c)"]),
|
||||
("not(a == 1)", [], [], []),
|
||||
("not(a == 1)", ["a"], [], ["not(a==1)"]),
|
||||
("a == 1 or b == 2", [], [], []),
|
||||
("a == 1 or b == 2", ["a"], [], ["a == 1"]),
|
||||
("a == 1 or b == 2", ["b"], [], ["b == 2"]),
|
||||
("a == 1 or b == 2", ["a", "b"], [], ["a == 1 or b == 2"]),
|
||||
("predicate(a,c) or predicate(b,c)", ["a", "b"], [], ["predicate(a,c) or predicate(b,c)"]),
|
||||
])
|
||||
def test_i_can_extract_predicates(expression, vars_to_include, vars_to_exclude, expected_expr):
|
||||
sheerka = get_sheerka()
|
||||
expected = [ast.parse(expr, mode="eval") for expr in expected_expr]
|
||||
|
||||
actual = core.builtin_helpers.extract_predicates(sheerka, expression, vars_to_include, vars_to_exclude)
|
||||
assert len(actual) == len(expected)
|
||||
for i in range(len(actual)):
|
||||
assert dump_ast(actual[i]) == dump_ast(expected[i])
|
||||
|
||||
|
||||
def get_sheerka():
|
||||
sheerka = Sheerka()
|
||||
sheerka = Sheerka(skip_builtins_in_db=True)
|
||||
sheerka.initialize("mem://")
|
||||
|
||||
return sheerka
|
||||
@@ -119,3 +162,10 @@ def get_sheerka():
|
||||
|
||||
def get_context(sheerka):
|
||||
return ExecutionContext("test", "xxx", sheerka)
|
||||
|
||||
|
||||
def dump_ast(node):
|
||||
dump = ast.dump(node)
|
||||
for to_remove in [", ctx=Load()", ", kind=None", ", type_ignores=[]"]:
|
||||
dump = dump.replace(to_remove, "")
|
||||
return dump
|
||||
|
||||
+23
-8
@@ -50,7 +50,7 @@ def test_i_can_list_builtin_concepts():
|
||||
|
||||
|
||||
def test_builtin_concepts_are_initialized():
|
||||
sheerka = get_sheerka()
|
||||
sheerka = get_sheerka(skip_builtins_in_db=False)
|
||||
assert len(sheerka.concepts_cache) == len(BuiltinConcepts)
|
||||
for concept_name in BuiltinConcepts:
|
||||
assert str(concept_name) in sheerka.concepts_cache
|
||||
@@ -61,7 +61,7 @@ def test_builtin_concepts_are_initialized():
|
||||
|
||||
|
||||
def test_builtin_concepts_can_be_updated():
|
||||
sheerka = get_sheerka(root_folder)
|
||||
sheerka = get_sheerka(root_folder, skip_builtins_in_db=False)
|
||||
loaded_sheerka = sheerka.get(BuiltinConcepts.SHEERKA)
|
||||
loaded_sheerka.desc = "I have a description"
|
||||
sheerka.sdp.modify("Test", sheerka.CONCEPTS_ENTRY, loaded_sheerka.key, loaded_sheerka)
|
||||
@@ -89,7 +89,8 @@ def test_i_can_add_a_concept():
|
||||
assert concept_found.id == "1001"
|
||||
|
||||
assert concept.key in sheerka.concepts_cache
|
||||
assert sheerka.sdp.io.exists(sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_found.get_digest()))
|
||||
assert sheerka.sdp.io.exists(
|
||||
sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_found.get_digest()))
|
||||
|
||||
|
||||
def test_i_cannot_add_the_same_concept_twice():
|
||||
@@ -414,7 +415,8 @@ as:
|
||||
assert getattr(concept_saved, prop) == getattr(expected, prop)
|
||||
|
||||
assert concept_saved.key in sheerka.concepts_cache
|
||||
assert sheerka.sdp.io.exists(sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_saved.get_digest()))
|
||||
assert sheerka.sdp.io.exists(
|
||||
sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_saved.get_digest()))
|
||||
|
||||
|
||||
def test_i_can_eval_def_concept_part_when_one_part_is_a_ref_of_another_concept():
|
||||
@@ -443,7 +445,8 @@ def test_i_can_eval_def_concept_part_when_one_part_is_a_ref_of_another_concept()
|
||||
assert getattr(concept_saved, prop) == getattr(expected, prop)
|
||||
|
||||
assert concept_saved.key in sheerka.concepts_cache
|
||||
assert sheerka.sdp.io.exists(sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_saved.get_digest()))
|
||||
assert sheerka.sdp.io.exists(
|
||||
sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_saved.get_digest()))
|
||||
|
||||
|
||||
def test_i_cannot_eval_the_same_def_concept_twice():
|
||||
@@ -551,12 +554,24 @@ def test_i_can_manage_concepts_with_the_same_key_when_values_are_the_same():
|
||||
res = sheerka.eval("hello 'foo'")
|
||||
assert len(res) == 1
|
||||
assert res[0].status
|
||||
assert res[0].value, "hello foo"
|
||||
assert res[0].value == "hello foo"
|
||||
assert res[0].who == sheerka.get_evaluator_name(MultipleSameSuccessEvaluator.NAME)
|
||||
|
||||
|
||||
def get_sheerka(root="mem://"):
|
||||
sheerka = Sheerka()
|
||||
def test_i_can_create_concepts_on_python_codes():
|
||||
sheerka = get_sheerka()
|
||||
context = get_context(sheerka)
|
||||
|
||||
sheerka.create_new_concept(context, Concept(name="concepts", body="sheerka.concepts()"))
|
||||
res = sheerka.eval("concepts")
|
||||
|
||||
assert len(res) == 1
|
||||
assert res[0].status
|
||||
assert isinstance(res[0].value, list)
|
||||
|
||||
|
||||
def get_sheerka(root="mem://", skip_builtins_in_db=True):
|
||||
sheerka = Sheerka(skip_builtins_in_db)
|
||||
sheerka.initialize(root)
|
||||
|
||||
return sheerka
|
||||
|
||||
@@ -51,8 +51,28 @@ def test_i_can_get_sub_classes():
|
||||
default_parser = core.utils.get_class("parsers.DefaultParser.DefaultParser")
|
||||
exact_concept_parser = core.utils.get_class("parsers.ExactConceptParser.ExactConceptParser")
|
||||
python_parser = core.utils.get_class("parsers.PythonParser.PythonParser")
|
||||
concept_lexer_parser = core.utils.get_class("parsers.ConceptLexerParser.ConceptLexerParser")
|
||||
|
||||
assert base_parser not in sub_classes
|
||||
assert default_parser in sub_classes
|
||||
assert exact_concept_parser in sub_classes
|
||||
assert python_parser in sub_classes
|
||||
assert concept_lexer_parser in sub_classes
|
||||
|
||||
|
||||
@pytest.mark.parametrize("a,b, expected", [
|
||||
([], [], []),
|
||||
([], ['a'], ['a']),
|
||||
([[]], ['a'], [['a']]),
|
||||
(['a'], [], ['a']),
|
||||
([['a']], [], [['a']]),
|
||||
|
||||
([['a']], ['b'], [['a', 'b']]),
|
||||
([['a'], ['b']], ['c'], [['a', 'c'], ['b', 'c']]),
|
||||
([['a1', 'a2'], ['b1', 'b2', 'b3']], ['c'], [['a1', 'a2', 'c'], ['b1', 'b2', 'b3', 'c']]),
|
||||
([[]], ['a', 'b'], [['a'], ['b']]),
|
||||
([['a'], ['b']], ['c', 'd', 'e'], [['a', 'c'], ['b', 'c'], ['a', 'd'], ['b', 'd'], ['a', 'e'], ['b', 'e']]),
|
||||
])
|
||||
def test_i_can_product(a, b, expected):
|
||||
res = core.utils.product(a, b)
|
||||
assert res == expected
|
||||
|
||||
Reference in New Issue
Block a user