pytrilogy 0.0.3.35__tar.gz → 0.0.3.37__tar.gz

{pytrilogy-0.0.3.35/pytrilogy.egg-info → pytrilogy-0.0.3.37}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.35
+Version: 0.0.3.37
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37/pytrilogy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.35
+Version: 0.0.3.37
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/tests/test_derived_concepts.py

@@ -11,6 +11,14 @@ def test_derivations(test_environment):
     )
 
 
+def test_hour_derivation(test_environment):
+    hour_derived = test_environment.concepts["order_timestamp.hour"]
+    assert "hour" in hour_derived.datatype.traits, hour_derived
+
+    year_derived = test_environment.concepts["order_timestamp.year"]
+    assert "year" in year_derived.datatype.traits, year_derived
+
+
 def test_filtering_where_on_derived_aggregate(test_environment):
     exception = False
     try:

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/tests/test_functions.py

@@ -212,6 +212,7 @@ def test_math_functions(test_environment):
     property order_add <- revenue + 2;
     property order_id.order_nested <- revenue * 2/2;
     property order_id.rounded <- round(revenue + 2.01,2);
+    constant random <- random(1);
     select
         order_id,
         inflated_order_value,
@@ -219,7 +220,8 @@ def test_math_functions(test_environment):
         fixed_order_value,
         order_sub,
         order_add,
-        rounded
+        rounded,
+        random,
     ;"""
     env, parsed = parse(declarations, environment=test_environment)
     select: SelectStatement = parsed[-1]
@@ -227,6 +229,22 @@ def test_math_functions(test_environment):
         dialect.compile_statement(process_query(test_environment, select))
 
 
+def test_random_randomness(test_environment):
+    declarations = """
+    auto x <- unnest([1,2,3,4]);
+    select
+        x,
+        random(x) -> random_1,
+    ;"""
+    env, parsed = parse(declarations, environment=test_environment)
+    z = (
+        Dialects.DUCK_DB.default_executor(environment=test_environment)
+        .execute_query(parsed[-1])
+        .fetchall()
+    )
+    assert z[0].random_1 != z[1].random_1, f"{z[0].random_1} == {z[1].random_1}"
+
+
 def test_string_functions(test_environment):
     declarations = """
     property test_name <- concat(category_name, '_test');

pytrilogy-0.0.3.37/tests/test_metadata.py

@@ -0,0 +1,26 @@
+from trilogy import parse
+
+
+def test_metadata():
+    env, _ = parse(
+        """key user_id int metadata(description="the description");
+property user_id.display_name string metadata(description="The display name");"""
+    )
+
+    assert env.concepts["user_id"].metadata.description == "the description"
+    assert env.concepts["display_name"].metadata.description == "The display name"
+
+
+# def test_import_metadata():
+#     env = Environment(working_path=Path(__file__).parent)
+#     env, _ = parse(
+#         """import test_env as env; # Dragon metrics
+#         import test_env as env2;""",
+#         environment=env,
+#     )
+
+#     assert "Dragon metrics" in env.concepts["env.id"].metadata.description
+
+#     env2 = env.concepts["env2.id"]
+#     assert env2.namespace == "env2"
+#     assert env.concepts["env2.id"].metadata.description is None

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/tests/test_typing.py

@@ -214,3 +214,24 @@ sum(
     )
 
     assert env.environment.concepts["total"].datatype.traits == ["money"]
+
+
+def test_custom_trait_unnest_typing():
+    env = Dialects.DUCK_DB.default_executor()
+    env.environment.parse(
+        """
+import std.geography;
+
+const array <- ['VT', 'MA', 'NY', 'CA']::array<string::us_state_short>;
+
+"""
+    )
+
+    _ = env.execute_query(
+        """
+SELECT
+    unnest(array)->state;
+"""
+    )
+
+    assert env.environment.concepts["state"].datatype.traits == ["us_state_short"]

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
 
-__version__ = "0.0.3.35"
+__version__ = "0.0.3.37"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/core/enums.py

@@ -155,6 +155,7 @@ class FunctionType(Enum):
     ROUND = "round"
     ABS = "abs"
     SQRT = "sqrt"
+    RANDOM = "random"
 
     # Aggregates
     ## group is not a real aggregate - it just means group by this + some other set of fields

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/core/environment_helpers.py

@@ -47,7 +47,7 @@ def generate_date_concepts(concept: Concept, environment: Environment):
         )
         new_concept = Concept(
             name=f"{concept.name}.{fname}",
-            datatype=DataType.INTEGER,
+            datatype=function.output_datatype,
             purpose=default_type,
             lineage=function,
             grain=concept.grain,
@@ -119,7 +119,7 @@ def generate_datetime_concepts(concept: Concept, environment: Environment):
         )
         new_concept = Concept(
             name=f"{concept.name}.{fname}",
-            datatype=DataType.INTEGER,
+            datatype=const_function.output_datatype,
             purpose=default_type,
             lineage=const_function,
             grain=concept.grain,

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/core/functions.py

@@ -620,6 +620,12 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         output_type=DataType.INTEGER,
         arg_count=1,
     ),
+    FunctionType.RANDOM: FunctionConfig(
+        valid_inputs=[],
+        output_purpose=Purpose.PROPERTY,
+        output_type=DataType.FLOAT,
+        arg_count=1,
+    ),
     FunctionType.ROUND: FunctionConfig(
         valid_inputs=[
             {DataType.INTEGER, DataType.FLOAT, DataType.NUMBER, DataType.NUMERIC},

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/core/models/build.py

@@ -1929,6 +1929,18 @@ class Factory:
     def _(self, base: TraitDataType):
         return base
 
+    @build.register
+    def _(self, base: ListType):
+        return base
+
+    @build.register
+    def _(self, base: StructType):
+        return base
+
+    @build.register
+    def _(self, base: MapType):
+        return base
+
     @build.register
     def _(self, base: Datasource):
         local_cache: dict[str, BuildConcept] = {}

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/dialect/base.py

@@ -132,6 +132,11 @@ DATATYPE_MAP: dict[DataType, str] = {
     DataType.MAP: "map",
     DataType.DATE: "date",
     DataType.DATETIME: "datetime",
+    DataType.LIST: "list",
+}
+
+COMPLEX_DATATYPE_MAP = {
+    DataType.LIST: lambda x: f"{x}[]",
 }
 
 
@@ -172,6 +177,7 @@ FUNCTION_MAP = {
     FunctionType.ROUND: lambda x: f"round({x[0]},{x[1]})",
     FunctionType.MOD: lambda x: f"({x[0]} % {x[1]})",
     FunctionType.SQRT: lambda x: f"sqrt({x[0]})",
+    FunctionType.RANDOM: lambda x: "random()",
     # aggregate types
     FunctionType.COUNT_DISTINCT: lambda x: f"count(distinct {x[0]})",
     FunctionType.COUNT: lambda x: f"count({x[0]})",
@@ -283,6 +289,7 @@ class BaseDialect:
     QUOTE_CHARACTER = "`"
     SQL_TEMPLATE = GENERIC_SQL_TEMPLATE
     DATATYPE_MAP = DATATYPE_MAP
+    COMPLEX_DATATYPE_MAP = COMPLEX_DATATYPE_MAP
     UNNEST_MODE = UnnestMode.CROSS_APPLY
 
     def __init__(self, rendering: Rendering | None = None):
@@ -682,6 +689,8 @@ class BaseDialect:
             return self.FUNCTION_MAP[FunctionType.DATETIME_LITERAL](e)
         elif isinstance(e, TraitDataType):
             return self.render_expr(e.type, cte=cte, cte_map=cte_map)
+        elif isinstance(e, ListType):
+            return f"{self.COMPLEX_DATATYPE_MAP[DataType.LIST](self.render_expr(e.value_data_type, cte=cte, cte_map=cte_map))}"
         else:
             raise ValueError(f"Unable to render type {type(e)} {e}")
 

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/dialect/bigquery.py

@@ -3,6 +3,9 @@ from typing import Any, Callable, Mapping
 from jinja2 import Template
 
 from trilogy.core.enums import FunctionType, UnnestMode, WindowType
+from trilogy.core.models.core import (
+    DataType,
+)
 from trilogy.dialect.base import BaseDialect
 
 WINDOW_FUNCTION_MAP: Mapping[WindowType, Callable[[Any, Any, Any], str]] = {}
@@ -36,6 +39,19 @@ FUNCTION_GRAIN_MATCH_MAP = {
     FunctionType.AVG: lambda args: f"{args[0]}",
 }
 
+DATATYPE_MAP: dict[DataType, str] = {
+    DataType.STRING: "STRING",
+    DataType.INTEGER: "INT64",
+    DataType.FLOAT: "FLOAT64",
+    DataType.BOOL: "BOOL",
+    DataType.NUMERIC: "NUMERIC",
+    DataType.MAP: "MAP",
+    DataType.DATE: "DATE",
+    DataType.DATETIME: "DATETIME",
+    DataType.TIMESTAMP: "TIMESTAMP",
+}
+
+
 BQ_SQL_TEMPLATE = Template(
     """{%- if output %}
 CREATE OR REPLACE TABLE {{ output.address.location }} AS
@@ -80,3 +96,4 @@ class BigqueryDialect(BaseDialect):
     QUOTE_CHARACTER = "`"
     SQL_TEMPLATE = BQ_SQL_TEMPLATE
     UNNEST_MODE = UnnestMode.CROSS_JOIN
+    DATATYPE_MAP = DATATYPE_MAP

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/parsing/parse_engine.py

@@ -368,18 +368,21 @@ class ParseToObjects(Transformer):
                     output.concept.metadata.description
                     or args[1].text.split("#")[1].strip()
                 )
-        if isinstance(output, ImportStatement):
-            if len(args) > 1 and isinstance(args[1], Comment):
-                comment = args[1].text.split("#")[1].strip()
-                namespace = output.alias
-                for _, v in self.environment.concepts.items():
-                    if v.namespace == namespace:
-                        if v.metadata.description:
-                            v.metadata.description = (
-                                f"{comment}: {v.metadata.description}"
-                            )
-                        else:
-                            v.metadata.description = comment
+        # this is a bad plan for now;
+        # because a comment after an import statement is very common
+        # and it's not intuitive that it modifies the import description
+        # if isinstance(output, ImportStatement):
+        #     if len(args) > 1 and isinstance(args[1], Comment):
+        #         comment = args[1].text.split("#")[1].strip()
+        #         namespace = output.alias
+        #         for _, v in self.environment.concepts.items():
+        #             if v.namespace == namespace:
+        #                 if v.metadata.description:
+        #                     v.metadata.description = (
+        #                         f"{comment}: {v.metadata.description}"
+        #                     )
+        #                 else:
+        #                     v.metadata.description = comment
 
         return args[0]
 
@@ -1762,6 +1765,10 @@ class ParseToObjects(Transformer):
     def fsqrt(self, meta: Meta, args) -> Function:
         return self.function_factory.create_function(args, FunctionType.SQRT, meta)
 
+    @v_args(meta=True)
+    def frandom(self, meta: Meta, args) -> Function:
+        return self.function_factory.create_function(args, FunctionType.RANDOM, meta)
+
     @v_args(meta=True)
     def fround(self, meta, args) -> Function:
         return self.function_factory.create_function(args, FunctionType.ROUND, meta)

{pytrilogy-0.0.3.35 → pytrilogy-0.0.3.37}/trilogy/parsing/trilogy.lark

@@ -196,9 +196,12 @@
     fmod: ( "mod"i "(" expr "," (int_lit | concept_lit ) ")") | ( expr "%" (int_lit | concept_lit ) )
     fround: "round"i "(" expr "," expr ")"
     fabs: "abs"i "(" expr ")"
-    fsqrt: "sqrt"i "(" expr ")"
+    _SQRT.1: "sqrt("
+    fsqrt:  _SQRT expr ")"
+    _RANDOM.1: "random("i
+    frandom: _RANDOM expr ")"
         
-    _math_functions: fmul | fdiv | fadd | fsub | fround | fmod | fabs | fsqrt
+    _math_functions: fmul | fdiv | fadd | fsub | fround | fmod | fabs | fsqrt | frandom
     
     //generic
     _fcast_primary: "cast"i "(" expr "as"i data_type ")"
@@ -354,7 +357,7 @@
 
     struct_type: "struct"i "<" ((data_type | IDENTIFIER) ",")* (data_type | IDENTIFIER) ","? ">" 
 
-    list_type: "list"i "<" data_type ">"
+    list_type: ("list"i "<" data_type ">" ) | ("array"i "<" data_type ">" )
 
     numeric_type: "numeric"i "(" int_lit "," int_lit ")"
 

pytrilogy-0.0.3.35/tests/test_metadata.py

@@ -1,29 +0,0 @@
-from pathlib import Path
-
-from trilogy import parse
-from trilogy.core.models.environment import Environment
-
-
-def test_metadata():
-    env, _ = parse(
-        """key user_id int metadata(description="the description");
-property user_id.display_name string metadata(description="The display name");"""
-    )
-
-    assert env.concepts["user_id"].metadata.description == "the description"
-    assert env.concepts["display_name"].metadata.description == "The display name"
-
-
-def test_import_metadata():
-    env = Environment(working_path=Path(__file__).parent)
-    env, _ = parse(
-        """import test_env as env; # Dragon metrics
-        import test_env as env2;""",
-        environment=env,
-    )
-
-    assert "Dragon metrics" in env.concepts["env.id"].metadata.description
-
-    env2 = env.concepts["env2.id"]
-    assert env2.namespace == "env2"
-    assert env.concepts["env2.id"].metadata.description is None