pytrilogy 0.0.2.49__tar.gz → 0.0.2.50__tar.gz

{pytrilogy-0.0.2.49/pytrilogy.egg-info → pytrilogy-0.0.2.50}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pytrilogy
-Version: 0.0.2.49
+Version: 0.0.2.50
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50/pytrilogy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pytrilogy
-Version: 0.0.2.49
+Version: 0.0.2.50
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/tests/test_discovery_nodes.py

@@ -15,7 +15,6 @@ def test_group_node(test_environment, test_environment_graph):
         output_concepts=[total_revenue, category],
         input_concepts=[category, revenue],
         environment=test_environment,
-        g=test_environment_graph,
         parents=[
             search_concepts(
                 [category, revenue],

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/tests/test_functions.py

@@ -14,6 +14,8 @@ from trilogy.dialect.duckdb import DuckDBDialect
 from trilogy.dialect.snowflake import SnowflakeDialect
 from trilogy.dialect.sql_server import SqlServerDialect
 from trilogy.parser import parse
+from trilogy import Dialects
+from datetime import date, datetime
 
 logger.setLevel(INFO)
 
@@ -154,6 +156,36 @@ def test_explicit_cast(test_environment):
         dialect.compile_statement(process_query(test_environment, select))
 
 
+def test_literal_cast(test_environment):
+    declarations = """
+select
+    '1'::int -> one,
+    '1'::float -> one_float,
+    '1'::string -> one_string,
+    '2024-01-01'::date -> one_date,
+    '2024-01-01 01:01:01'::datetime -> one_datetime,
+    'true'::bool -> one_bool,
+    ;"""
+    env, parsed = parse(declarations, environment=test_environment)
+
+    select: SelectStatement = parsed[-1]
+    z = (
+        Dialects.DUCK_DB.default_executor(environment=test_environment)
+        .execute_query(parsed[-1])
+        .fetchall()
+    )
+    assert z[0].one == 1
+    assert z[0].one_float == 1.0
+    assert z[0].one_string == "1"
+    assert z[0].one_date == date(year=2024, month=1, day=1)
+    assert z[0].one_datetime == datetime(
+        year=2024, month=1, day=1, hour=1, minute=1, second=1
+    )
+    assert z[0].one_bool == True
+    for dialect in TEST_DIALECTS:
+        dialect.compile_statement(process_query(test_environment, select))
+
+
 def test_math_functions(test_environment):
     declarations = """
     

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
 
-__version__ = "0.0.2.49"
+__version__ = "0.0.2.50"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/enums.py

@@ -120,6 +120,8 @@ class FunctionType(Enum):
 
     ALIAS = "alias"
 
+    PARENTHETICAL = "parenthetical"
+
     # Generic
     CASE = "case"
     CAST = "cast"
@@ -135,6 +137,8 @@ class FunctionType(Enum):
     ATTR_ACCESS = "attr_access"
     STRUCT = "struct"
     ARRAY = "array"
+    DATE_LITERAL = "date_literal"
+    DATETIME_LITERAL = "datetime_literal"
 
     # TEXT AND MAYBE MORE
     SPLIT = "split"
@@ -260,6 +264,13 @@ class ComparisonOperator(Enum):
     CONTAINS = "contains"
     ELSE = "else"
 
+    def __eq__(self, other):
+        if isinstance(other, str):
+            return self.value == other
+        if not isinstance(other, ComparisonOperator):
+            return False
+        return self.value == other.value
+
     @classmethod
     def _missing_(cls, value):
         if not isinstance(value, list) and " " in str(value):

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/functions.py

@@ -1,3 +1,4 @@
+from datetime import date, datetime
 from typing import Optional
 
 from trilogy.constants import MagicConstants
@@ -17,6 +18,8 @@ from trilogy.core.models import (
     arg_to_datatype,
 )
 
+GENERIC_ARGS = Concept | Function | str | int | float | date | datetime
+
 
 def create_function_derived_concept(
     name: str,
@@ -262,7 +265,7 @@ def get_attr_datatype(
     return arg.datatype
 
 
-def AttrAccess(args: list[Concept]):
+def AttrAccess(args: list[GENERIC_ARGS]):
     return Function(
         operator=FunctionType.ATTR_ACCESS,
         arguments=args,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/models.py

@@ -5,6 +5,7 @@ import hashlib
 import os
 from abc import ABC
 from collections import UserDict, UserList, defaultdict
+from datetime import date, datetime
 from enum import Enum
 from functools import cached_property
 from pathlib import Path
@@ -1264,6 +1265,8 @@ class Function(Mergeable, Namespaced, SelectContext, BaseModel):
             int,
             float,
             str,
+            date,
+            datetime,
             MapWrapper[Any, Any],
             DataType,
             ListType,
@@ -3868,6 +3871,8 @@ class Comparison(
         float,
         list,
         bool,
+        datetime,
+        date,
         Function,
         Concept,
         "Conditional",
@@ -3884,6 +3889,8 @@ class Comparison(
         float,
         list,
         bool,
+        date,
+        datetime,
         Concept,
         Function,
         "Conditional",
@@ -5008,5 +5015,9 @@ def arg_to_datatype(arg) -> DataType | ListType | StructType | MapType | Numeric
         return ListType(type=wrapper.type)
     elif isinstance(arg, MapWrapper):
         return MapType(key_type=arg.key_type, value_type=arg.value_type)
+    elif isinstance(arg, datetime):
+        return DataType.DATETIME
+    elif isinstance(arg, date):
+        return DataType.DATE
     else:
         raise ValueError(f"Cannot parse arg datatype for arg of raw type {type(arg)}")

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/concept_strategies_v3.py

@@ -359,7 +359,6 @@ def generate_node(
             input_concepts=[],
             output_concepts=constant_targets,
             environment=environment,
-            g=g,
             parents=[],
             depth=depth + 1,
         )
@@ -906,7 +905,6 @@ def _search_concepts(
                 input_concepts=non_virtual,
                 output_concepts=non_virtual,
                 environment=environment,
-                g=g,
                 parents=stack,
                 depth=depth,
             )
@@ -987,7 +985,6 @@ def source_query_concepts(
             x for x in root.output_concepts if x.address not in root.hidden_concepts
         ],
         environment=environment,
-        g=g,
         parents=[root],
         partial_concepts=root.partial_concepts,
     )

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/common.py

@@ -130,7 +130,6 @@ def gen_property_enrichment_node(
         ),
         output_concepts=base_node.output_concepts + extra_properties,
         environment=environment,
-        g=g,
         parents=[
             base_node,
         ]
@@ -209,7 +208,6 @@ def gen_enrichment_node(
         input_concepts=unique(join_keys + extra_required + non_hidden, "address"),
         output_concepts=unique(join_keys + extra_required + non_hidden, "address"),
         environment=environment,
-        g=g,
         parents=[enrich_node, base_node],
         force_group=False,
         preexisting_conditions=conditions.conditional if conditions else None,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/filter_node.py

@@ -117,7 +117,6 @@ def gen_filter_node(
                 input_concepts=row_parent.output_concepts,
                 output_concepts=[concept] + row_parent.output_concepts,
                 environment=row_parent.environment,
-                g=row_parent.g,
                 parents=[row_parent],
                 depth=row_parent.depth,
                 partial_concepts=row_parent.partial_concepts,
@@ -161,7 +160,6 @@ def gen_filter_node(
             ),
             output_concepts=[concept, immediate_parent] + parent_row_concepts,
             environment=environment,
-            g=g,
             parents=core_parents,
             grain=Grain(
                 components=[immediate_parent] + parent_row_concepts,
@@ -202,7 +200,6 @@ def gen_filter_node(
         ]
         + local_optional,
         environment=environment,
-        g=g,
         parents=[
             # this node fetches only what we need to filter
             filter_node,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/group_node.py

@@ -100,7 +100,6 @@ def gen_group_node(
         output_concepts=output_concepts,
         input_concepts=parent_concepts,
         environment=environment,
-        g=g,
         parents=parents,
         depth=depth,
         preexisting_conditions=conditions.conditional if conditions else None,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/group_to_node.py

@@ -45,7 +45,6 @@ def gen_group_to_node(
         output_concepts=parent_concepts + [concept],
         input_concepts=parent_concepts,
         environment=environment,
-        g=g,
         parents=parents,
         depth=depth,
     )
@@ -76,7 +75,6 @@ def gen_group_to_node(
         + [x for x in parent_concepts if x.address != concept.address],
         output_concepts=[concept] + local_optional,
         environment=environment,
-        g=g,
         parents=[
             # this node gets the group
             group_node,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/multiselect_node.py

@@ -108,7 +108,6 @@ def gen_multiselect_node(
         input_concepts=[x for y in base_parents for x in y.output_concepts],
         output_concepts=[x for y in base_parents for x in y.output_concepts],
         environment=environment,
-        g=g,
         depth=depth,
         parents=base_parents,
         node_joins=node_joins,
@@ -178,7 +177,6 @@ def gen_multiselect_node(
         input_concepts=enrich_node.output_concepts + node.output_concepts,
         output_concepts=node.output_concepts + local_optional,
         environment=environment,
-        g=g,
         depth=depth,
         parents=[
             # this node gets the multiselect

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/node_merge_node.py

@@ -333,7 +333,6 @@ def subgraphs_to_merge_node(
         input_concepts=unique(input_c, "address"),
         output_concepts=[x for x in all_concepts],
         environment=environment,
-        g=g,
         parents=parents,
         depth=depth,
         # conditions=conditions,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/rowset_node.py

@@ -140,7 +140,6 @@ def gen_rowset_node(
         input_concepts=non_hidden + non_hidden_enrich,
         output_concepts=non_hidden + local_optional,
         environment=environment,
-        g=g,
         depth=depth,
         parents=[
             node,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/select_merge_node.py

@@ -13,6 +13,9 @@ from trilogy.core.models import (
     LooseConceptList,
     WhereClause,
 )
+from trilogy.core.processing.node_generators.select_helpers.datasource_injection import (
+    get_union_sources,
+)
 from trilogy.core.processing.nodes import (
     ConstantNode,
     GroupNode,
@@ -35,38 +38,66 @@ def extract_address(node: str):
 def get_graph_partial_nodes(
     g: nx.DiGraph, conditions: WhereClause | None
 ) -> dict[str, list[str]]:
-    datasources: dict[str, Datasource] = nx.get_node_attributes(g, "datasource")
+    datasources: dict[str, Datasource | list[Datasource]] = nx.get_node_attributes(
+        g, "datasource"
+    )
     partial: dict[str, list[str]] = {}
     for node in g.nodes:
         if node in datasources:
             ds = datasources[node]
-            partial[node] = [concept_to_node(c) for c in ds.partial_concepts]
-            if ds.non_partial_for and conditions == ds.non_partial_for:
+            if not isinstance(ds, list):
+                if ds.non_partial_for and conditions == ds.non_partial_for:
+                    partial[node] = []
+                    continue
+                partial[node] = [concept_to_node(c) for c in ds.partial_concepts]
+                ds = [ds]
+            # assume union sources have no partial
+            else:
                 partial[node] = []
 
     return partial
 
 
 def get_graph_grain_length(g: nx.DiGraph) -> dict[str, int]:
-    datasources: dict[str, Datasource] = nx.get_node_attributes(g, "datasource")
-    partial: dict[str, int] = {}
+    datasources: dict[str, Datasource | list[Datasource]] = nx.get_node_attributes(
+        g, "datasource"
+    )
+    grain_length: dict[str, int] = {}
     for node in g.nodes:
         if node in datasources:
-            partial[node] = len(datasources[node].grain.components)
-    return partial
+            lookup = datasources[node]
+            if not isinstance(lookup, list):
+                lookup = [lookup]
+            assert isinstance(lookup, list)
+            grain_length[node] = sum(len(x.grain.components) for x in lookup)
+    return grain_length
 
 
 def create_pruned_concept_graph(
     g: nx.DiGraph,
     all_concepts: List[Concept],
+    datasources: list[Datasource],
     accept_partial: bool = False,
     conditions: WhereClause | None = None,
 ) -> nx.DiGraph:
     orig_g = g
     g = g.copy()
+
+    union_options = get_union_sources(datasources, all_concepts)
+    for ds_list in union_options:
+        node_address = "ds~" + "-".join([x.name for x in ds_list])
+        common: set[Concept] = set.intersection(
+            *[set(x.output_concepts) for x in ds_list]
+        )
+        g.add_node(node_address, datasource=ds_list)
+        for c in common:
+            g.add_edge(node_address, concept_to_node(c))
+
     target_addresses = set([c.address for c in all_concepts])
     concepts: dict[str, Concept] = nx.get_node_attributes(orig_g, "concept")
-    datasources: dict[str, Datasource] = nx.get_node_attributes(orig_g, "datasource")
+    datasource_map: dict[str, Datasource | list[Datasource]] = nx.get_node_attributes(
+        orig_g, "datasource"
+    )
     relevant_concepts_pre = {
         n: x.address
         for n in g.nodes()
@@ -81,13 +112,13 @@ def create_pruned_concept_graph(
         to_remove = []
         for edge in g.edges:
             if (
-                edge[0] in datasources
+                edge[0] in datasource_map
                 and (pnodes := partial.get(edge[0], []))
                 and edge[1] in pnodes
             ):
                 to_remove.append(edge)
             if (
-                edge[1] in datasources
+                edge[1] in datasource_map
                 and (pnodes := partial.get(edge[1], []))
                 and edge[0] in pnodes
             ):
@@ -136,7 +167,9 @@ def create_pruned_concept_graph(
     for edge in orig_g.edges():
         if edge[0] in relevant and edge[1] in relevant:
             g.add_edge(edge[0], edge[1])
-
+    # if we have no ds nodes at all, for non constant, we can't find it
+    if not any([n.startswith("ds~") for n in g.nodes]):
+        return None
     return g
 
 
@@ -190,6 +223,54 @@ def resolve_subgraphs(
     return pruned_subgraphs
 
 
+def create_datasource_node(
+    datasource: Datasource,
+    all_concepts: List[Concept],
+    accept_partial: bool,
+    environment: Environment,
+    depth: int,
+    conditions: WhereClause | None = None,
+) -> tuple[StrategyNode, bool]:
+    target_grain = Grain(components=all_concepts)
+    force_group = False
+    if not datasource.grain.issubset(target_grain):
+        force_group = True
+    partial_concepts = [
+        c.concept
+        for c in datasource.columns
+        if not c.is_complete and c.concept.address in all_concepts
+    ]
+    partial_lcl = LooseConceptList(concepts=partial_concepts)
+    nullable_concepts = [
+        c.concept
+        for c in datasource.columns
+        if c.is_nullable and c.concept.address in all_concepts
+    ]
+    nullable_lcl = LooseConceptList(concepts=nullable_concepts)
+    partial_is_full = conditions and (conditions == datasource.non_partial_for)
+    return (
+        SelectNode(
+            input_concepts=[c.concept for c in datasource.columns],
+            output_concepts=all_concepts,
+            environment=environment,
+            parents=[],
+            depth=depth,
+            partial_concepts=(
+                [] if partial_is_full else [c for c in all_concepts if c in partial_lcl]
+            ),
+            nullable_concepts=[c for c in all_concepts if c in nullable_lcl],
+            accept_partial=accept_partial,
+            datasource=datasource,
+            grain=Grain(components=all_concepts),
+            conditions=datasource.where.conditional if datasource.where else None,
+            preexisting_conditions=(
+                conditions.conditional if partial_is_full and conditions else None
+            ),
+        ),
+        force_group,
+    )
+
+
 def create_select_node(
     ds_name: str,
     subgraph: list[str],
@@ -199,12 +280,11 @@ def create_select_node(
     depth: int,
     conditions: WhereClause | None = None,
 ) -> StrategyNode:
-    ds_name = ds_name.split("~")[1]
+
     all_concepts = [
         environment.concepts[extract_address(c)] for c in subgraph if c.startswith("c~")
     ]
 
-    all_lcl = LooseConceptList(concepts=all_concepts)
     if all([c.derivation == PurposeLineage.CONSTANT for c in all_concepts]):
         logger.info(
             f"{padding(depth)}{LOGGER_PREFIX} All concepts {[x.address for x in all_concepts]} are constants, returning constant node"
@@ -213,7 +293,6 @@ def create_select_node(
             output_concepts=all_concepts,
             input_concepts=[],
             environment=environment,
-            g=g,
             parents=[],
             depth=depth,
             # no partial for constants
@@ -221,41 +300,44 @@ def create_select_node(
             force_group=False,
         )
 
-    datasource = environment.datasources[ds_name]
-    target_grain = Grain(components=all_concepts)
-    force_group = False
-    if not datasource.grain.issubset(target_grain):
-        force_group = True
-    partial_concepts = [
-        c.concept
-        for c in datasource.columns
-        if not c.is_complete and c.concept in all_lcl
-    ]
-    partial_lcl = LooseConceptList(concepts=partial_concepts)
-    nullable_concepts = [
-        c.concept for c in datasource.columns if c.is_nullable and c.concept in all_lcl
-    ]
-    nullable_lcl = LooseConceptList(concepts=nullable_concepts)
-    partial_is_full = conditions and (conditions == datasource.non_partial_for)
-    bcandidate: StrategyNode = SelectNode(
-        input_concepts=[c.concept for c in datasource.columns],
-        output_concepts=all_concepts,
-        environment=environment,
-        g=g,
-        parents=[],
-        depth=depth,
-        partial_concepts=(
-            [] if partial_is_full else [c for c in all_concepts if c in partial_lcl]
-        ),
-        nullable_concepts=[c for c in all_concepts if c in nullable_lcl],
-        accept_partial=accept_partial,
-        datasource=datasource,
-        grain=Grain(components=all_concepts),
-        conditions=datasource.where.conditional if datasource.where else None,
-        preexisting_conditions=(
-            conditions.conditional if partial_is_full and conditions else None
-        ),
-    )
+    datasource: dict[str, Datasource | list[Datasource]] = nx.get_node_attributes(
+        g, "datasource"
+    )[ds_name]
+    if isinstance(datasource, Datasource):
+        bcandidate, force_group = create_datasource_node(
+            datasource,
+            all_concepts,
+            accept_partial,
+            environment,
+            depth,
+            conditions=conditions,
+        )
+
+    elif isinstance(datasource, list):
+        from trilogy.core.processing.nodes.union_node import UnionNode
+
+        force_group = False
+        parents = []
+        for x in datasource:
+            subnode, fg = create_datasource_node(
+                x,
+                all_concepts,
+                accept_partial,
+                environment,
+                depth,
+                conditions=conditions,
+            )
+            parents.append(subnode)
+            force_group = force_group or fg
+        bcandidate = UnionNode(
+            output_concepts=all_concepts,
+            input_concepts=all_concepts,
+            environment=environment,
+            parents=parents,
+            depth=depth,
+        )
+    else:
+        raise ValueError(f"Unknown datasource type {datasource}")
 
     # we need to nest the group node one further
     if force_group is True:
@@ -263,14 +345,11 @@ def create_select_node(
             output_concepts=all_concepts,
             input_concepts=all_concepts,
             environment=environment,
-            g=g,
             parents=[bcandidate],
             depth=depth,
             partial_concepts=bcandidate.partial_concepts,
             nullable_concepts=bcandidate.nullable_concepts,
-            preexisting_conditions=(
-                conditions.conditional if partial_is_full and conditions else None
-            ),
+            preexisting_conditions=bcandidate.preexisting_conditions,
         )
     else:
         candidate = bcandidate
@@ -292,7 +371,6 @@ def gen_select_merge_node(
             output_concepts=constants,
             input_concepts=[],
             environment=environment,
-            g=g,
             parents=[],
             depth=depth,
             partial_concepts=[],
@@ -300,7 +378,11 @@ def gen_select_merge_node(
         )
     for attempt in [False, True]:
         pruned_concept_graph = create_pruned_concept_graph(
-            g, non_constant, attempt, conditions
+            g,
+            non_constant,
+            accept_partial=attempt,
+            conditions=conditions,
+            datasources=list(environment.datasources.values()),
         )
         if pruned_concept_graph:
             logger.info(
@@ -321,7 +403,7 @@ def gen_select_merge_node(
         create_select_node(
             k,
             subgraph,
-            g=g,
+            g=pruned_concept_graph,
             accept_partial=accept_partial,
             environment=environment,
             depth=depth,
@@ -338,7 +420,6 @@ def gen_select_merge_node(
                 output_concepts=constants,
                 input_concepts=[],
                 environment=environment,
-                g=g,
                 parents=[],
                 depth=depth,
                 partial_concepts=[],
@@ -361,7 +442,6 @@ def gen_select_merge_node(
         output_concepts=all_concepts,
         input_concepts=non_constant,
         environment=environment,
-        g=g,
         depth=depth,
         parents=parents,
         preexisting_conditions=preexisting_conditions,
@@ -372,7 +452,6 @@ def gen_select_merge_node(
             output_concepts=all_concepts,
             input_concepts=all_concepts,
             environment=environment,
-            g=g,
             parents=[base],
             depth=depth,
             preexisting_conditions=preexisting_conditions,

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/union_node.py

@@ -70,6 +70,5 @@ def gen_union_node(
         input_concepts=[concept] + local_optional,
         output_concepts=[concept] + local_optional,
         environment=environment,
-        g=g,
         parents=parents,
     )

{pytrilogy-0.0.2.49 → pytrilogy-0.0.2.50}/trilogy/core/processing/node_generators/unnest_node.py

@@ -46,7 +46,6 @@ def gen_unnest_node(
         input_concepts=arguments + non_equivalent_optional,
         output_concepts=[concept] + local_optional,
         environment=environment,
-        g=g,
         parents=([parent] if (arguments or local_optional) else []),
     )
     # we need to sometimes nest an unnest node,
@@ -56,7 +55,6 @@ def gen_unnest_node(
         input_concepts=base.output_concepts,
         output_concepts=base.output_concepts,
         environment=environment,
-        g=g,
         parents=[base],
         preexisting_conditions=conditions.conditional if conditions else None,
     )