pytrilogy 0.0.3.112__tar.gz → 0.0.3.115__tar.gz

{pytrilogy-0.0.3.112/pytrilogy.egg-info → pytrilogy-0.0.3.115}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.112
+Version: 0.0.3.115
 Summary: Declarative, typed query language that compiles to SQL.
 Classifier: Programming Language :: Python
 Classifier: Programming Language :: Python :: 3
@@ -51,61 +51,41 @@ The Trilogy language is an experiment in better SQL for analytics - a streamline
 
 Trilogy is especially powerful for data consumption, providing a rich metadata layer that makes creating, interpreting, and visualizing queries easy and expressive.
 
+
+We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
+
+
 ## Quick Start
 
 > [!TIP]
 > **Try it now:** [Open-source studio](https://trilogydata.dev/trilogy-studio-core/) | [Interactive demo](https://trilogydata.dev/demo/) | [Documentation](https://trilogydata.dev/)
 
-**Install locally:**
+**Install**
 ```bash
 pip install pytrilogy
 ```
 
-**Your first query:**
+**Save in hello.preql**
 ```sql
-# Save as hello.preql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
+const prime <- unnest([2, 3, 5, 7, 11, 13, 17, 19, 23, 29]);
 
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
+def cube_plus_one(x) -> (x * x * x + 1);
 
 WHERE 
-    @initcap(name) in top_names
+    prime_cubed_plus_one % 7 = 0
 SELECT
-    name,
-    sum(births) as name_count
+    prime,
+    @cube_plus_one(prime) as prime_cubed_plus_one
 ORDER BY
-    name_count desc
+    prime asc
 LIMIT 10;
 ```
 
-**Run it:**
+**Run it in DuckDB**
 ```bash
 trilogy run hello.preql duckdb
 ```
 
-We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
-
-## Trilogy Looks Like SQL
-
-```sql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
-
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
-
-WHERE 
-    @initcap(name) in top_names
-SELECT
-    name,
-    sum(births) as name_count
-ORDER BY
-    name_count desc
-LIMIT 10;
-```
-
 ## Trilogy is Easy to Write
 For humans *and* AI. Enjoy flexible, one-shot query generation without any DB access or security risks. 
 

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/README.md

@@ -19,61 +19,41 @@ The Trilogy language is an experiment in better SQL for analytics - a streamline
 
 Trilogy is especially powerful for data consumption, providing a rich metadata layer that makes creating, interpreting, and visualizing queries easy and expressive.
 
+
+We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
+
+
 ## Quick Start
 
 > [!TIP]
 > **Try it now:** [Open-source studio](https://trilogydata.dev/trilogy-studio-core/) | [Interactive demo](https://trilogydata.dev/demo/) | [Documentation](https://trilogydata.dev/)
 
-**Install locally:**
+**Install**
 ```bash
 pip install pytrilogy
 ```
 
-**Your first query:**
+**Save in hello.preql**
 ```sql
-# Save as hello.preql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
+const prime <- unnest([2, 3, 5, 7, 11, 13, 17, 19, 23, 29]);
 
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
+def cube_plus_one(x) -> (x * x * x + 1);
 
 WHERE 
-    @initcap(name) in top_names
+    prime_cubed_plus_one % 7 = 0
 SELECT
-    name,
-    sum(births) as name_count
+    prime,
+    @cube_plus_one(prime) as prime_cubed_plus_one
 ORDER BY
-    name_count desc
+    prime asc
 LIMIT 10;
 ```
 
-**Run it:**
+**Run it in DuckDB**
 ```bash
 trilogy run hello.preql duckdb
 ```
 
-We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
-
-## Trilogy Looks Like SQL
-
-```sql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
-
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
-
-WHERE 
-    @initcap(name) in top_names
-SELECT
-    name,
-    sum(births) as name_count
-ORDER BY
-    name_count desc
-LIMIT 10;
-```
-
 ## Trilogy is Easy to Write
 For humans *and* AI. Enjoy flexible, one-shot query generation without any DB access or security risks. 
 

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115/pytrilogy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.112
+Version: 0.0.3.115
 Summary: Declarative, typed query language that compiles to SQL.
 Classifier: Programming Language :: Python
 Classifier: Programming Language :: Python :: 3
@@ -51,61 +51,41 @@ The Trilogy language is an experiment in better SQL for analytics - a streamline
 
 Trilogy is especially powerful for data consumption, providing a rich metadata layer that makes creating, interpreting, and visualizing queries easy and expressive.
 
+
+We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
+
+
 ## Quick Start
 
 > [!TIP]
 > **Try it now:** [Open-source studio](https://trilogydata.dev/trilogy-studio-core/) | [Interactive demo](https://trilogydata.dev/demo/) | [Documentation](https://trilogydata.dev/)
 
-**Install locally:**
+**Install**
 ```bash
 pip install pytrilogy
 ```
 
-**Your first query:**
+**Save in hello.preql**
 ```sql
-# Save as hello.preql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
+const prime <- unnest([2, 3, 5, 7, 11, 13, 17, 19, 23, 29]);
 
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
+def cube_plus_one(x) -> (x * x * x + 1);
 
 WHERE 
-    @initcap(name) in top_names
+    prime_cubed_plus_one % 7 = 0
 SELECT
-    name,
-    sum(births) as name_count
+    prime,
+    @cube_plus_one(prime) as prime_cubed_plus_one
 ORDER BY
-    name_count desc
+    prime asc
 LIMIT 10;
 ```
 
-**Run it:**
+**Run it in DuckDB**
 ```bash
 trilogy run hello.preql duckdb
 ```
 
-We recommend starting with the studio to explore Trilogy. For integration, `pytrilogy` can be run locally to parse and execute trilogy model [.preql] files using the `trilogy` CLI tool, or can be run in python by importing the `trilogy` package.
-
-## Trilogy Looks Like SQL
-
-```sql
-import names;
-
-const top_names <- ['Elvis', 'Elvira', 'Elrond', 'Sam'];
-
-def initcap(word) -> upper(substring(word, 1, 1)) || substring(word, 2, len(word));
-
-WHERE 
-    @initcap(name) in top_names
-SELECT
-    name,
-    sum(births) as name_count
-ORDER BY
-    name_count desc
-LIMIT 10;
-```
-
 ## Trilogy is Easy to Write
 For humans *and* AI. Enjoy flexible, one-shot query generation without any DB access or security risks. 
 

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/tests/test_functions.py

@@ -146,6 +146,8 @@ def test_date_functions(test_environment):
         date_trunc(order_timestamp, hour) -> order_hour_trunc,
         date_trunc(order_timestamp, minute) -> order_minute_trunc,
         date_trunc(order_timestamp, second) -> order_second_trunc,
+        date_trunc(order_timestamp, quarter) -> order_quarter_trunc,
+        date_trunc(order_timestamp, week) -> order_week_trunc,
         date_part(order_timestamp, month) -> order_month_part,
         date_part(order_timestamp, day) -> order_day_part,
         date_part(order_timestamp, year) -> order_year_part,
@@ -155,6 +157,8 @@ def test_date_functions(test_environment):
         date_part(order_timestamp, quarter) -> order_quarter_part,
         date_part(order_timestamp, week) -> order_week_part,
         date_part(order_timestamp, day_of_week) -> order_day_of_week_part,
+        month_name(order_timestamp) -> order_month_name,
+        day_name(order_timestamp) -> order_day_name
     ;
     
     
@@ -163,7 +167,28 @@ def test_date_functions(test_environment):
     select: SelectStatement = parsed[-1]
 
     for dialect in TEST_DIALECTS:
+        engine = None
+        if isinstance(dialect, BigqueryDialect):
+            engine = Dialects.DUCK_DB.default_executor(environment=test_environment)
         dialect.compile_statement(process_query(test_environment, select))
+        if engine:
+            engine.execute_raw_sql(
+                """CREATE TABLE tblRevenue AS
+SELECT 
+    100.50 as revenue,
+    1 as order_id,
+    101 as product_id,
+    TIMESTAMP '2024-01-15 10:30:00' as order_timestamp
+UNION ALL
+SELECT 
+    250.75 as revenue,
+    2 as order_id,
+    102 as product_id,
+    TIMESTAMP '2024-01-16 14:45:30' as order_timestamp;"""
+            )
+            results = engine.execute_query(select)
+            assert results
+            assert results.fetchall()
 
 
 def test_bad_cast(test_environment):

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/tests/test_parse_engine.py

@@ -121,3 +121,17 @@ def test_alias_error():
     with raises(InvalidSyntaxException) as e:
         env.parse(TEXT2)
     assert ERROR_CODES[201] in str(e.value), e.value
+
+
+def test_semicolon_error():
+    env = Environment()
+    TEXT2 = """
+    const a <- 1;
+
+    select
+        a+2 as fun,
+    
+    """
+    with raises(InvalidSyntaxException) as e:
+        env.parse(TEXT2)
+    assert ERROR_CODES[202] in str(e.value), e.value

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
 
-__version__ = "0.0.3.112"
+__version__ = "0.0.3.115"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/constants.py

@@ -1,7 +1,9 @@
 import random
+from contextlib import contextmanager
 from dataclasses import dataclass, field
 from enum import Enum
 from logging import getLogger
+from typing import Any
 
 logger = getLogger("trilogy")
 
@@ -50,6 +52,32 @@ class Rendering:
     parameters: bool = True
     concise: bool = False
 
+    @contextmanager
+    def temporary(self, **kwargs: Any):
+        """
+        Context manager to temporarily set attributes and revert them afterwards.
+
+        Usage:
+            r = Rendering()
+            with r.temporary(parameters=False, concise=True):
+                # parameters is False, concise is True here
+                do_something()
+            # parameters and concise are back to their original values
+        """
+        # Store original values
+        original_values = {key: getattr(self, key) for key in kwargs}
+
+        # Set new values
+        for key, value in kwargs.items():
+            setattr(self, key, value)
+
+        try:
+            yield self
+        finally:
+            # Restore original values
+            for key, value in original_values.items():
+                setattr(self, key, value)
+
 
 @dataclass
 class Parsing:

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/enums.py

@@ -169,6 +169,7 @@ class FunctionType(Enum):
     ARRAY_SORT = "array_sort"
     ARRAY_TRANSFORM = "array_transform"
     ARRAY_TO_STRING = "array_to_string"
+    ARRAY_FILTER = "array_filter"
 
     # MAP
     MAP_KEYS = "map_keys"
@@ -204,6 +205,7 @@ class FunctionType(Enum):
     MIN = "min"
     AVG = "avg"
     ARRAY_AGG = "array_agg"
+    ANY = "any"
 
     # String
     LIKE = "like"
@@ -236,12 +238,15 @@ class FunctionType(Enum):
     MONTH = "month"
     QUARTER = "quarter"
     YEAR = "year"
+    MONTH_NAME = "month_name"
+    DAY_NAME = "day_name"
 
     DATE_PART = "date_part"
     DATE_TRUNCATE = "date_truncate"
     DATE_ADD = "date_add"
     DATE_SUB = "date_sub"
     DATE_DIFF = "date_diff"
+    DATE_SPINE = "date_spine"
 
     # UNIX
     UNIX_TO_TIMESTAMP = "unix_to_timestamp"
@@ -261,6 +266,7 @@ class FunctionClass(Enum):
         FunctionType.ARRAY_AGG,
         FunctionType.COUNT,
         FunctionType.COUNT_DISTINCT,
+        FunctionType.ANY,
     ]
     SINGLE_ROW = [
         FunctionType.CONSTANT,
@@ -268,7 +274,7 @@ class FunctionClass(Enum):
         FunctionType.CURRENT_DATETIME,
     ]
 
-    ONE_TO_MANY = [FunctionType.UNNEST]
+    ONE_TO_MANY = [FunctionType.UNNEST, FunctionType.DATE_SPINE]
 
     RECURSIVE = [FunctionType.RECURSE_EDGE]
 

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/functions.py

@@ -175,6 +175,10 @@ def get_date_trunc_output(
         return DataType.DATETIME
     elif target == DatePart.SECOND:
         return DataType.DATETIME
+    elif target == DatePart.WEEK:
+        return DataType.DATE
+    elif target == DatePart.QUARTER:
+        return DataType.DATE
     else:
         raise InvalidSyntaxException(f"Date truncation not supported for {target}")
 
@@ -208,6 +212,14 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         output_type_function=get_unnest_output_type,
         arg_count=1,
     ),
+    FunctionType.DATE_SPINE: FunctionConfig(
+        valid_inputs={
+            DataType.DATE,
+        },
+        output_purpose=Purpose.KEY,
+        output_type=DataType.DATE,
+        arg_count=2,
+    ),
     FunctionType.RECURSE_EDGE: FunctionConfig(
         arg_count=2,
     ),
@@ -314,6 +326,18 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         output_type_function=get_transform_output_type,
         arg_count=3,
     ),
+    FunctionType.ARRAY_FILTER: FunctionConfig(
+        valid_inputs=[
+            {
+                DataType.ARRAY,
+            },
+            {*DataType},
+            {*DataType},
+        ],
+        output_purpose=Purpose.PROPERTY,
+        output_type_function=get_transform_output_type,
+        arg_count=3,
+    ),
     FunctionType.ARRAY_TO_STRING: FunctionConfig(
         valid_inputs={
             DataType.ARRAY,
@@ -640,6 +664,17 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         output_type=TraitDataType(type=DataType.INTEGER, traits=["day"]),
         arg_count=1,
     ),
+    FunctionType.DAY_NAME: FunctionConfig(
+        valid_inputs={
+            DataType.DATE,
+            DataType.TIMESTAMP,
+            DataType.DATETIME,
+            # DataType.STRING,
+        },
+        output_purpose=Purpose.PROPERTY,
+        output_type=TraitDataType(type=DataType.STRING, traits=["day_name"]),
+        arg_count=1,
+    ),
     FunctionType.WEEK: FunctionConfig(
         valid_inputs={
             DataType.DATE,
@@ -662,6 +697,17 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         output_type=TraitDataType(type=DataType.INTEGER, traits=["month"]),
         arg_count=1,
     ),
+    FunctionType.MONTH_NAME: FunctionConfig(
+        valid_inputs={
+            DataType.DATE,
+            DataType.TIMESTAMP,
+            DataType.DATETIME,
+            # DataType.STRING,
+        },
+        output_purpose=Purpose.PROPERTY,
+        output_type=TraitDataType(type=DataType.STRING, traits=["month_name"]),
+        arg_count=1,
+    ),
     FunctionType.QUARTER: FunctionConfig(
         valid_inputs={
             DataType.DATE,
@@ -861,6 +907,11 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         ),
         arg_count=1,
     ),
+    FunctionType.ANY: FunctionConfig(
+        valid_inputs={*DataType},
+        output_purpose=Purpose.PROPERTY,
+        arg_count=1,
+    ),
     FunctionType.AVG: FunctionConfig(
         valid_inputs={
             DataType.INTEGER,

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/models/author.py

@@ -461,6 +461,11 @@ class WhereClause(Mergeable, ConceptArgs, Namespaced, BaseModel):
             conditional=self.conditional.with_namespace(namespace)
         )
 
+    def with_reference_replacement(self, source, target):
+        return self.__class__.model_construct(
+            conditional=self.conditional.with_reference_replacement(source, target)
+        )
+
 
 class HavingClause(WhereClause):
     pass
@@ -1212,6 +1217,8 @@ class Concept(Addressable, DataTyped, ConceptArgs, Mergeable, Namespaced, BaseMo
             return Derivation.FILTER
         elif lineage and isinstance(lineage, (BuildAggregateWrapper, AggregateWrapper)):
             return Derivation.AGGREGATE
+        # elif lineage and isinstance(lineage, (BuildParenthetical, Parenthetical)):
+        #     return Derivation.PARENTHETICAL
         elif lineage and isinstance(lineage, (BuildRowsetItem, RowsetItem)):
             return Derivation.ROWSET
         elif lineage and isinstance(
@@ -1227,7 +1234,7 @@ class Concept(Addressable, DataTyped, ConceptArgs, Mergeable, Namespaced, BaseMo
         elif (
             lineage
             and isinstance(lineage, (BuildFunction, Function))
-            and lineage.operator == FunctionType.UNNEST
+            and lineage.operator in FunctionClass.ONE_TO_MANY.value
         ):
             return Derivation.UNNEST
         elif (
@@ -1279,7 +1286,8 @@ class Concept(Addressable, DataTyped, ConceptArgs, Mergeable, Namespaced, BaseMo
         elif (
             lineage
             and isinstance(lineage, (Function, BuildFunction))
-            and lineage.operator in (FunctionType.UNNEST, FunctionType.UNION)
+            and lineage.operator
+            in (FunctionType.UNNEST, FunctionType.UNION, FunctionType.DATE_SPINE)
         ):
             return Granularity.MULTI_ROW
         elif lineage and all(

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/models/build.py

@@ -134,8 +134,9 @@ def concept_is_relevant(
     if concept.purpose in (Purpose.METRIC,):
         if all([c in others for c in concept.grain.components]):
             return False
+    if concept.derivation in (Derivation.UNNEST,):
+        return True
     if concept.derivation in (Derivation.BASIC,):
-
         return any(concept_is_relevant(c, others) for c in concept.concept_arguments)
     if concept.granularity == Granularity.SINGLE_ROW:
         return False
@@ -1668,7 +1669,6 @@ class Factory:
                     valid_inputs=base.valid_inputs,
                     arg_count=base.arg_count,
                 )
-
         new = BuildFunction(
             operator=base.operator,
             arguments=[self.handle_constant(self.build(c)) for c in raw_args],
@@ -1724,6 +1724,14 @@ class Factory:
         return self._build_concept(base)
 
     def _build_concept(self, base: Concept) -> BuildConcept:
+        try:
+            return self.__build_concept(base)
+        except RecursionError as e:
+            raise RecursionError(
+                f"Recursion error building concept {base.address}. This is likely due to a circular reference."
+            ) from e
+
+    def __build_concept(self, base: Concept) -> BuildConcept:
         # TODO: if we are using parameters, wrap it in a new model and use that in rendering
         if base.address in self.local_concepts:
             return self.local_concepts[base.address]
@@ -2002,6 +2010,13 @@ class Factory:
     def _build_tuple_wrapper(self, base: TupleWrapper) -> TupleWrapper:
         return TupleWrapper(val=[self.build(x) for x in base.val], type=base.type)
 
+    @build.register
+    def _(self, base: ListWrapper) -> ListWrapper:
+        return self._build_list_wrapper(base)
+
+    def _build_list_wrapper(self, base: ListWrapper) -> ListWrapper:
+        return ListWrapper([self.build(x) for x in base], type=base.type)
+
     @build.register
     def _(self, base: FilterItem) -> BuildFilterItem:
         return self._build_filter_item(base)

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/models/environment.py

@@ -563,6 +563,7 @@ class Environment(BaseModel):
             existing = self.validate_concept(concept, meta=meta)
             if existing:
                 concept = existing
+
         self.concepts[concept.address] = concept
 
         from trilogy.core.environment_helpers import generate_related_concepts

{pytrilogy-0.0.3.112 → pytrilogy-0.0.3.115}/trilogy/core/processing/concept_strategies_v3.py

@@ -306,7 +306,12 @@ def evaluate_loop_conditions(
 
 
 def check_for_early_exit(
-    complete, partial, missing, context: LoopContext, priority_concept: BuildConcept
+    complete: ValidationResult,
+    found: set[str],
+    partial: set[str],
+    missing: set[str],
+    context: LoopContext,
+    priority_concept: BuildConcept,
 ) -> bool:
     if complete == ValidationResult.INCOMPLETE_CONDITION:
         cond_dict = {str(node): node.preexisting_conditions for node in context.stack}
@@ -331,8 +336,18 @@ def check_for_early_exit(
                 f"{depth_to_prefix(context.depth)}{LOGGER_PREFIX} Breaking as we have attempted all nodes"
             )
             return True
+        elif all(
+            [
+                x.address in found and x.address not in partial
+                for x in context.mandatory_list
+            ]
+        ):
+            logger.info(
+                f"{depth_to_prefix(context.depth)}{LOGGER_PREFIX} Breaking as we have found all mandatory nodes without partials"
+            )
+            return True
         logger.info(
-            f"{depth_to_prefix(context.depth)}{LOGGER_PREFIX} Found complete stack with partials {partial}, continuing search, attempted {context.attempted} all {len(context.mandatory_list)}"
+            f"{depth_to_prefix(context.depth)}{LOGGER_PREFIX} Found complete stack with partials {partial}, continuing search, attempted {context.attempted} of total {len(context.mandatory_list)}."
         )
     else:
         logger.info(
@@ -436,6 +451,7 @@ def generate_loop_completion(context: LoopContext, virtual: set[str]) -> Strateg
             context.original_mandatory,
             context.environment,
             non_virtual_difference_values,
+            depth=context.depth,
         )
 
     return group_if_required_v2(
@@ -443,6 +459,7 @@ def generate_loop_completion(context: LoopContext, virtual: set[str]) -> Strateg
         context.original_mandatory,
         context.environment,
         non_virtual_difference_values,
+        depth=context.depth,
     )
 
 
@@ -466,6 +483,7 @@ def _search_concepts(
         conditions=conditions,
     )
 
+    # if we get a can
     if candidate:
         return candidate
     context = initialize_loop_context(
@@ -477,13 +495,14 @@ def _search_concepts(
         accept_partial=accept_partial,
         conditions=conditions,
     )
-
+    partial: set[str] = set()
     while context.incomplete:
 
         priority_concept = get_priority_concept(
             context.mandatory_list,
             context.attempted,
             found_concepts=context.found,
+            partial_concepts=partial,
             depth=depth,
         )
 
@@ -538,7 +557,7 @@ def _search_concepts(
         # assign
         context.found = found_c
         early_exit = check_for_early_exit(
-            complete, partial, missing_c, context, priority_concept
+            complete, found_c, partial, missing_c, context, priority_concept
         )
         if early_exit:
             break
@@ -608,4 +627,4 @@ def source_query_concepts(
     logger.info(
         f"{depth_to_prefix(0)}{LOGGER_PREFIX} final concepts are {[x.address for x in final]}"
     )
-    return group_if_required_v2(root, output_concepts, environment)
+    return group_if_required_v2(root, output_concepts, environment, depth=0)