pytrilogy 0.0.1.117__tar.gz → 0.0.2.1__tar.gz

{pytrilogy-0.0.1.117/pytrilogy.egg-info → pytrilogy-0.0.2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pytrilogy
-Version: 0.0.1.117
+Version: 0.0.2.1
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1/pytrilogy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pytrilogy
-Version: 0.0.1.117
+Version: 0.0.2.1
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/pytrilogy.egg-info/SOURCES.txt

@@ -58,7 +58,6 @@ trilogy/core/processing/utility.py
 trilogy/core/processing/node_generators/__init__.py
 trilogy/core/processing/node_generators/basic_node.py
 trilogy/core/processing/node_generators/common.py
-trilogy/core/processing/node_generators/concept_merge_node.py
 trilogy/core/processing/node_generators/filter_node.py
 trilogy/core/processing/node_generators/group_node.py
 trilogy/core/processing/node_generators/group_to_node.py

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/tests/test_discovery_nodes.py

@@ -39,7 +39,11 @@ def test_group_node_property(test_environment: Environment, test_environment_gra
         source_concepts=search_concepts,
         depth=0,
     )
-    input_concept_names = {x.name for x in group_node.parents[0].output_concepts}
+    input_concept_names = {
+        x.name
+        for x in group_node.parents[0].output_concepts
+        if x not in group_node.parents[0].hidden_concepts
+    }
     assert input_concept_names == {"category_name_length", "category_id"}
     # assert len(input_concept.grain.components) == 1
     # assert input_concept.grain.components[0].name == "category_id"
@@ -60,7 +64,11 @@ def test_group_node_property_all(test_environment: Environment, test_environment
         source_concepts=search_concepts,
         depth=0,
     )
-    input_concept_names = {x.name for x in group_node.parents[0].output_concepts}
+    input_concept_names = {
+        x.name
+        for x in group_node.parents[0].output_concepts
+        if x not in group_node.parents[0].hidden_concepts
+    }
     assert input_concept_names == {"category_name_length", "category_id"}
     final = group_node.resolve()
     assert len(final.datasources) == 1

pytrilogy-0.0.2.1/tests/test_environment.py

@@ -0,0 +1,64 @@
+from trilogy.core.models import Environment
+from pathlib import Path
+from trilogy.core.enums import Modifier
+
+
+def test_environment_serialization(test_environment: Environment):
+    str(test_environment)
+    path = test_environment.to_cache()
+
+    test_environment2 = Environment.from_cache(path)
+    assert test_environment2
+
+    assert test_environment.concepts == test_environment2.concepts
+    assert test_environment.datasources == test_environment2.datasources
+    for k, v in test_environment.concepts.items():
+        assert v == test_environment2.concepts[k]
+
+
+def test_environment_from_path():
+
+    env = Environment.from_file(Path(__file__).parent / "test_env.preql")
+
+    assert "id" in env.concepts
+
+
+def test_environment_merge():
+    env1: Environment
+    env1, _ = Environment().parse(
+        """
+key  order_id int;   
+                                
+                                datasource orders
+                                (order_id:order_id)
+                                grain (order_id)
+                                address orders;
+"""
+    )
+
+    env2, _ = Environment().parse(
+        """
+                                key order_id int;
+                                                                
+                                datasource replacements
+                                (order_id:order_id)
+                                grain (order_id)
+                                address replacements;
+
+                                
+"""
+    )
+
+    env1.add_import("replacements", env2)
+
+    _ = env1.merge_concept(
+        env1.concepts["replacements.order_id"],
+        env1.concepts["order_id"],
+        modifiers=[Modifier.PARTIAL],
+    )
+
+    assert env1.concepts["order_id"] == env1.concepts["replacements.order_id"]
+
+    order_id = env1.datasources["replacements.replacements"].columns[0]
+    assert order_id.concept == env1.concepts["order_id"]
+    assert order_id.modifiers == [Modifier.PARTIAL]

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/tests/test_models.py

@@ -132,7 +132,7 @@ def test_undefined(test_environment: Environment):
         environment=test_environment.concepts,
     )
 
-    y = x.with_select_grain(Grain(components=[test_environment.concepts["order_id"]]))
+    y = x.with_select_context(Grain(components=[test_environment.concepts["order_id"]]))
 
     assert y.grain == Grain(components=[test_environment.concepts["order_id"]])
 

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/tests/test_parsing.py

@@ -15,6 +15,7 @@ from trilogy.parsing.parse_engine import (
 )
 from trilogy.constants import MagicConstants
 from trilogy.dialect.base import BaseDialect
+from trilogy.core.enums import BooleanOperator
 
 
 def test_in():
@@ -127,6 +128,15 @@ def test_show(test_environment):
     )
 
 
+def test_conditional(test_environment):
+    _, parsed = parse_text(
+        "const order_id <- 4; SELECT order_id  WHERE order_id =4 and order_id = 10;"
+    )
+    query = parsed[-1]
+    assert isinstance(query, SelectStatement)
+    assert query.where_clause.conditional.operator == BooleanOperator.AND
+
+
 def test_as_transform(test_environment):
     _, parsed = parse_text("const order_id <- 4; SELECT order_id as new_order_id;")
     query = parsed[-1]

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/tests/test_query_processing.py

@@ -134,7 +134,7 @@ def test_query_aggregation(test_environment, test_environment_graph):
         environment=test_environment, graph=test_environment_graph, statement=select
     )
 
-    assert {datasource.identifier} == {"revenue_at_local_order_id_at_abstract"}
+    assert {datasource.identifier} == {"local_revenue_at_local_order_id_at_abstract"}
     check = datasource
     assert len(check.input_concepts) == 2
     assert check.input_concepts[0].name == "revenue"

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.executor import Executor
 from trilogy.parser import parse
 from trilogy.constants import CONFIG
 
-__version__ = "0.0.1.117"
+__version__ = "0.0.2.1"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/trilogy/constants.py

@@ -1,6 +1,7 @@
 from logging import getLogger
 from dataclasses import dataclass, field
 from enum import Enum
+import random
 
 logger = getLogger("trilogy")
 
@@ -34,7 +35,12 @@ class Config:
     validate_missing: bool = True
     optimizations: Optimizations = field(default_factory=Optimizations)
 
+    def set_random_seed(self, seed: int):
+        random.seed(seed)
+
 
 CONFIG = Config()
 
+CONFIG.set_random_seed(42)
+
 CONFIG.strict_mode = True

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/trilogy/core/enums.py

@@ -43,7 +43,6 @@ class PurposeLineage(Enum):
     ROOT = "root"
     ROWSET = "rowset"
     MULTISELECT = "multiselect"
-    MERGE = "merge"
 
 
 class Granularity(Enum):
@@ -263,7 +262,6 @@ class DatePart(Enum):
 class SourceType(Enum):
     FILTER = "filter"
     SELECT = "select"
-    MERGE = "merge"
     ABSTRACT = "abstract"
     DIRECT_SELECT = "direct_select"
     GROUP = "group"
@@ -271,8 +269,15 @@ class SourceType(Enum):
     UNNEST = "unnest"
     CONSTANT = "constant"
     ROWSET = "rowset"
+    MERGE = "merge"
 
 
 class ShowCategory(Enum):
     MODELS = "models"
     CONCEPTS = "concepts"
+
+
+class SelectFiltering(Enum):
+    NONE = "none"
+    EXPLICIT = "explicit"  # the filtering contains only selected values
+    IMPLICIT = "implicit"  # the filtering contains unselected values

pytrilogy-0.0.2.1/trilogy/core/env_processor.py

@@ -0,0 +1,63 @@
+from trilogy.core.graph_models import (
+    ReferenceGraph,
+    concept_to_node,
+    datasource_to_node,
+)
+from trilogy.core.models import Environment, Concept, Datasource
+
+
+def add_concept(concept: Concept, g: ReferenceGraph):
+    g.add_node(concept)
+    # if we have sources, recursively add them
+    node_name = concept_to_node(concept)
+    if concept.sources:
+        for source in concept.sources:
+            generic = source.with_default_grain()
+            g.add_edge(generic, node_name)
+    for _, pseudonym in concept.pseudonyms.items():
+        pseudonym = pseudonym.with_default_grain()
+        pseudonym_node = concept_to_node(pseudonym)
+        if pseudonym_node.split("@")[0] == node_name.split("@")[0]:
+            continue
+        g.add_edge(pseudonym_node, node_name, pseudonym=True)
+        g.add_edge(node_name, pseudonym_node, pseudonym=True)
+        add_concept(pseudonym, g)
+
+
+def generate_adhoc_graph(
+    concepts: list[Concept],
+    datasources: list[Datasource],
+    restrict_to_listed: bool = False,
+) -> ReferenceGraph:
+    g = ReferenceGraph()
+
+    # add all parsed concepts
+    for concept in concepts:
+        add_concept(concept, g)
+
+    for dataset in datasources:
+        node = datasource_to_node(dataset)
+        g.add_node(dataset, type="datasource", datasource=dataset)
+        for concept in dataset.concepts:
+            if restrict_to_listed:
+                if concept_to_node(concept) not in g.nodes:
+                    continue
+            g.add_edge(node, concept)
+            g.add_edge(concept, node)
+            # if there is a key on a table at a different grain
+            # add an FK edge to the canonical source, if it exists
+            # for example, order ID on order product table
+            default = concept.with_default_grain()
+            if concept != default:
+                g.add_edge(concept, default)
+                g.add_edge(default, concept)
+    return g
+
+
+def generate_graph(
+    environment: Environment,
+) -> ReferenceGraph:
+
+    return generate_adhoc_graph(
+        list(environment.concepts.values()), list(environment.datasources.values())
+    )

{pytrilogy-0.0.1.117 → pytrilogy-0.0.2.1}/trilogy/core/functions.py

@@ -52,12 +52,22 @@ def argument_to_purpose(arg) -> Purpose:
     if isinstance(arg, Function):
         return arg.output_purpose
     elif isinstance(arg, AggregateWrapper):
+        base = arg.function.output_purpose
+        if arg.by and base == Purpose.METRIC:
+            return Purpose.PROPERTY
         return arg.function.output_purpose
     elif isinstance(arg, Parenthetical):
         return argument_to_purpose(arg.content)
     elif isinstance(arg, WindowItem):
         return Purpose.PROPERTY
     elif isinstance(arg, Concept):
+        base = arg.purpose
+        if (
+            isinstance(arg.lineage, AggregateWrapper)
+            and arg.lineage.by
+            and base == Purpose.METRIC
+        ):
+            return Purpose.PROPERTY
         return arg.purpose
     elif isinstance(arg, (int, float, str, bool, list, NumericType, DataType)):
         return Purpose.CONSTANT
@@ -153,6 +163,7 @@ def Max(args: list[Concept]) -> Function:
             DataType.DATE,
             DataType.DATETIME,
             DataType.TIMESTAMP,
+            DataType.BOOL,
         },
         arg_count=1,
         # output_grain=Grain(components=arguments),