Refactored Caching, Refactored BnfNodeParser, Introduced Sphinx

This commit is contained in:
2020-05-12 17:21:10 +02:00
parent 7d3a490bc5
commit 6e343ba996
110 changed files with 13865 additions and 7540 deletions
+95 -150
View File
@@ -1,18 +1,20 @@
from core.builtin_concepts import BuiltinConcepts
from core.concept import PROPERTIES_TO_SERIALIZE, Concept, DEFINITION_TYPE_DEF
from core.sheerka.Sheerka import Sheerka
from sdp.sheerkaDataProvider import SheerkaDataProvider
from tests.TestUsingFileBasedSheerka import TestUsingFileBasedSheerka
from tests.TestUsingMemoryBasedSheerka import TestUsingMemoryBasedSheerka
class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
def test_i_can_add_a_concept(self):
sheerka = self.get_sheerka()
def test_i_can_create_a_concept(self):
sheerka = self.get_sheerka(cache_only=False)
context = self.get_context(sheerka)
concept = self.get_default_concept()
res = sheerka.create_new_concept(self.get_context(sheerka), concept)
res = sheerka.create_new_concept(context, concept)
sheerka.cache_manager.commit(context)
assert res.status
assert sheerka.isinstance(res.value, BuiltinConcepts.NEW_CONCEPT)
@@ -24,22 +26,36 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
assert concept_found.key == "__var__0 + __var__1"
assert concept_found.id == "1001"
assert concept.key in sheerka.cache_by_key
assert concept.id in sheerka.cache_by_id
assert concept.name in sheerka.cache_by_name
assert sheerka.sdp.io.exists(
sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_found.get_origin()))
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, concept.get_definition_hash())
# saved in cache
assert sheerka.has_id(concept.id)
assert sheerka.has_key(concept.key)
assert sheerka.has_name(concept.name)
assert sheerka.has_hash(concept.get_definition_hash())
# I can get the concept using various index
assert sheerka.get_by_id(concept.id) == concept
assert sheerka.get_by_key(concept.key) == concept
assert sheerka.get_by_name(concept.name) == concept
assert sheerka.get_by_hash(concept.get_definition_hash()) == concept
# I can get by the first entry
assert sheerka.cache_manager.get(sheerka.CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "+") == [concept.id]
assert sheerka.cache_manager.get(sheerka.RESOLVED_CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "+") == [concept.id]
# saved in sdp
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_ID_ENTRY, concept.id)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_KEY_ENTRY, concept.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_NAME_ENTRY, concept.name)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_ENTRY, concept.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, concept.get_definition_hash())
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "+")
def test_i_can_add_a_concept_when_name_differs_from_the_key(self):
sheerka = self.get_sheerka()
concept = Concept("greetings", definition="hello a", definition_type=DEFINITION_TYPE_DEF).def_prop("a")
sheerka = self.get_sheerka(cache_only=False)
context = self.get_context(sheerka)
concept = Concept("greetings", definition="hello a", definition_type=DEFINITION_TYPE_DEF).def_var("a")
res = sheerka.create_new_concept(self.get_context(sheerka), concept)
sheerka.cache_manager.commit(context)
assert res.status
assert sheerka.isinstance(res.value, BuiltinConcepts.NEW_CONCEPT)
@@ -51,15 +67,24 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
assert concept_found.key == "hello __var__0"
assert concept_found.id == "1001"
assert concept.key in sheerka.cache_by_key
assert concept.id in sheerka.cache_by_id
assert concept.name in sheerka.cache_by_name
assert sheerka.sdp.io.exists(
sheerka.sdp.io.get_obj_path(SheerkaDataProvider.ObjectsFolder, concept_found.get_origin()))
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, concept.get_definition_hash())
# saved in cache
assert sheerka.has_id(concept.id)
assert sheerka.has_key(concept.key)
assert sheerka.has_name(concept.name)
assert sheerka.has_hash(concept.get_definition_hash())
# I can get the concept using various index
assert sheerka.get_by_id(concept.id) == concept
assert sheerka.get_by_key(concept.key) == concept
assert sheerka.get_by_name(concept.name) == concept
assert sheerka.get_by_hash(concept.get_definition_hash()) == concept
# saved in sdp
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_ID_ENTRY, concept.id)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_KEY_ENTRY, concept.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_NAME_ENTRY, concept.name)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_ENTRY, concept.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, concept.get_definition_hash())
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "hello")
def test_i_cannot_add_the_same_concept_twice(self):
"""
@@ -82,7 +107,7 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
sheerka.create_new_concept(self.get_context(sheerka), concept)
from_cache = sheerka.get(concept.key)
from_cache = sheerka.get_by_key(concept.key)
assert from_cache is not None
assert from_cache == concept
@@ -90,57 +115,7 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
assert from_cache is not None
assert from_cache == concept
def test_i_first_look_in_local_cache(self):
sheerka = self.get_sheerka()
concept = self.get_default_concept()
sheerka.create_new_concept(self.get_context(sheerka), concept)
sheerka.cache_by_key[concept.key].pre = "I have modified the concept in cache"
from_cache = sheerka.get(concept.key)
assert from_cache is not None
assert from_cache.key == concept.key
assert from_cache.pre == "I have modified the concept in cache"
def test_i_can_get_a_known_concept_when_not_in_cache(self):
"""
When not in cache, uses sdp
:return:
"""
sheerka = self.get_sheerka()
concept = self.get_default_concept()
sheerka.create_new_concept(self.get_context(sheerka), concept)
sheerka.reset_cache()
loaded = sheerka.get(concept.key)
assert loaded == concept
# I can also get it by its id
loaded = sheerka.sdp.get(sheerka.CONCEPTS_BY_ID_ENTRY, concept.id)
assert loaded == concept
def test_i_can_instantiate_a_concept_from_sdp(self):
sheerka = self.get_sheerka()
concept = Concept("foo")
sheerka.create_new_concept(self.get_context(sheerka), concept)
sheerka.reset_cache()
loaded = sheerka.new("foo")
assert loaded == concept
def test_i_can_get_a_concept_by_its_id(self):
sheerka = self.get_sheerka()
concept = self.get_default_concept()
sheerka.create_new_concept(self.get_context(sheerka), concept)
sheerka.cache_by_key = {} # reset the cache
loaded = sheerka.get_by_id(concept.id)
assert loaded == concept
def test_i_can_get_list_of_concept_when_same_key_when_no_cache(self):
def test_i_can_get_list_of_concept_when_same_key_using_cache(self):
sheerka = self.get_sheerka()
concept1 = self.get_default_concept()
concept2 = self.get_default_concept()
@@ -149,79 +124,13 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
res1 = sheerka.create_new_concept(self.get_context(sheerka), concept1)
res2 = sheerka.create_new_concept(self.get_context(sheerka), concept2)
assert res1.status
assert res2.status
assert res1.value.body.key == res2.value.body.key # same key
sheerka.cache_by_key = {} # reset the cache
result = sheerka.get(concept1.key)
result = sheerka.get_by_key(concept1.key)
assert len(result) == 2
assert result[0] == concept1
assert result[1] == concept2
def test_i_can_get_list_of_concept_when_same_key_when_cache(self):
sheerka = self.get_sheerka()
concept1 = self.get_default_concept()
concept2 = self.get_default_concept()
concept2.metadata.body = "a+b"
res1 = sheerka.create_new_concept(self.get_context(sheerka), concept1)
res2 = sheerka.create_new_concept(self.get_context(sheerka), concept2)
assert res1.value.body.key == res2.value.body.key # same key
# sheerka.cache_by_key = {} # Do not reset the cache
result = sheerka.get(concept1.key)
assert len(result) == 2
assert result[0] == concept1
assert result[1] == concept2
def test_i_can_get_the_correct_concept_using_the_id_when_same_key_when_no_cache(self):
sheerka = self.get_sheerka()
concept1 = self.get_default_concept()
concept2 = self.get_default_concept()
concept2.metadata.body = "a+b"
res1 = sheerka.create_new_concept(self.get_context(sheerka), concept1)
res2 = sheerka.create_new_concept(self.get_context(sheerka), concept2)
assert res1.value.body.key == res2.value.body.key # same key
result = sheerka.get(concept1.key, res2.body.body.id)
assert result.name == "a + b"
assert result.metadata.body == "a+b"
def test_i_can_get_the_correct_concept_using_the_id__when_same_key_when_cache(self):
sheerka = self.get_sheerka()
concept1 = self.get_default_concept()
concept2 = self.get_default_concept()
concept2.metadata.body = "a+b"
res1 = sheerka.create_new_concept(self.get_context(sheerka), concept1)
res2 = sheerka.create_new_concept(self.get_context(sheerka), concept2)
assert res1.value.body.key == res2.value.body.key # same key
result = sheerka.get(concept1.key, res2.body.body.id)
assert result.name == "a + b"
assert result.metadata.body == "a+b"
def test_i_cannot_get_the_correct_concept_id_the_id_is_wrong(self):
sheerka = self.get_sheerka()
concept1 = self.get_default_concept()
concept2 = self.get_default_concept()
concept2.metadata.body = "a+b"
res1 = sheerka.create_new_concept(self.get_context(sheerka), concept1)
res2 = sheerka.create_new_concept(self.get_context(sheerka), concept2)
assert res1.value.body.key == res2.value.body.key # same key
result = sheerka.get(concept1.key, "wrong id")
assert sheerka.isinstance(result, BuiltinConcepts.UNKNOWN_CONCEPT)
def test_concept_that_references_itself_is_correctly_created(self):
sheerka = self.get_sheerka()
concept = Concept("foo", body="foo")
@@ -233,27 +142,63 @@ class TestSheerkaCreateNewConcept(TestUsingMemoryBasedSheerka):
class TestSheerkaCreateNewConceptFileBased(TestUsingFileBasedSheerka):
def test_i_can_add_several_concepts(self):
sheerka, context, hello, greeting = self.init_concepts(
Concept("Hello world a").def_prop("a"),
Concept("Greeting a").def_prop("a"),
use_dict=False
)
res = sheerka.create_new_concept(self.get_context(sheerka), hello)
sheerka = self.get_sheerka()
context = self.get_context(sheerka)
hello = Concept("Hello world a").def_var("a")
res = sheerka.create_new_concept(context, hello)
sheerka.cache_manager.commit(context)
assert res.status
sheerka = self.get_sheerka(use_dict=False)
res = sheerka.create_new_concept(self.get_context(sheerka), greeting)
sheerka = self.get_sheerka() # another instance
context = self.get_context(sheerka)
greeting = Concept("Greeting a").def_var("a")
res = sheerka.create_new_concept(context, greeting)
sheerka.cache_manager.commit(context)
assert res.status
assert sheerka.sdp.exists(Sheerka.CONCEPTS_ENTRY, hello.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_ENTRY, greeting.key)
sheerka = self.get_sheerka() # another instance again
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_KEY_ENTRY, hello.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_KEY_ENTRY, greeting.key)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_ID_ENTRY, hello.id)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_ID_ENTRY, greeting.id)
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_NAME_ENTRY, "Hello world a")
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_NAME_ENTRY, "Greeting a")
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, hello.get_definition_hash())
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_HASH_ENTRY, greeting.get_definition_hash())
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "Hello")
assert sheerka.sdp.exists(Sheerka.CONCEPTS_BY_FIRST_KEYWORD_ENTRY, "Greeting")
def test_i_cannot_add_the_same_concept_twice_using_sdp(self):
"""
Checks that duplicated concepts are managed by sheerka, not by sheerka.sdp
:return:
"""
sheerka = self.get_sheerka(cache_only=False)
context = self.get_context(sheerka)
concept = self.get_default_concept()
sheerka.create_new_concept(context, concept)
sheerka.cache_manager.commit(context)
sheerka.cache_manager.clear()
res = sheerka.create_new_concept(context, concept)
assert not res.status
assert sheerka.isinstance(res.value, BuiltinConcepts.CONCEPT_ALREADY_DEFINED)
assert res.value.body == concept
def test_new_entry_does_not_override_the_previous_ones(self):
sheerka = self.get_sheerka()
context = self.get_context(sheerka)
sheerka.create_new_concept(context, Concept("foo", body="1"))
sheerka.create_new_concept(context, Concept("foo", body="2"))
sheerka.cache_manager.commit(context)
assert len(sheerka.sdp.get(Sheerka.CONCEPTS_BY_KEY_ENTRY, "foo")) == 2
sheerka = self.get_sheerka() # new instance
sheerka.create_new_concept(context, Concept("foo", body="3"))
sheerka.cache_manager.commit(context)
assert len(sheerka.sdp.get(Sheerka.CONCEPTS_BY_KEY_ENTRY, "foo")) == 3