pytrilogy 0.0.3.6__py3-none-any.whl → 0.0.3.8__py3-none-any.whl

{pytrilogy-0.0.3.6.dist-info → pytrilogy-0.0.3.8.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: pytrilogy
-Version: 0.0.3.6
+Version: 0.0.3.8
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.3.6.dist-info → pytrilogy-0.0.3.8.dist-info}/RECORD

@@ -1,8 +1,8 @@
-trilogy/__init__.py,sha256=R9yJSDmZQvpxjWMcp4mGFnbg7xuxUCiIVrvP8eacKj4,302
+trilogy/__init__.py,sha256=OtihFBeuWR7p7GBhiNMfyCpmu-x2kn_W2xSawos5ILM,302
 trilogy/compiler.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/constants.py,sha256=qZ1d0hoKPPV2HHCoFwPYTVB7b6bXjpWvXd3lE-zEhy8,1494
-trilogy/engine.py,sha256=yOPnR7XCjWG82Gym_LLZBkYKKJdLCvqdCyt8zguNcnM,1103
-trilogy/executor.py,sha256=sssEPDnIDPiQtMSrt5pFiJXUfcDc6gSi4m2Eliod_BM,16844
+trilogy/engine.py,sha256=3etkm2RSVKO0IkgPKkrcs33X5gN_fIMyqMNfChcsR1E,1318
+trilogy/executor.py,sha256=YgSCeeYVecI9526LGSLVe2apOo7Ddsttvs_nDC9yElQ,17194
 trilogy/parser.py,sha256=o4cfk3j3yhUFoiDKq9ZX_GjBF3dKhDjXEwb63rcBkBM,293
 trilogy/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/utility.py,sha256=euQccZLKoYBz0LNg5tzLlvv2YHvXh9HArnYp1V3uXsM,763
@@ -14,18 +14,18 @@ trilogy/core/env_processor.py,sha256=pFsxnluKIusGKx1z7tTnfsd_xZcPy9pZDungkjkyvI0
 trilogy/core/environment_helpers.py,sha256=oOpewPwMp8xOtx2ayeeyuLNUwr-cli7UanHKot5ebNY,7627
 trilogy/core/ergonomics.py,sha256=ASLDd0RqKWrZiG3XcKHo8nyTjaB_8xfE9t4NZ1UvGpc,1639
 trilogy/core/exceptions.py,sha256=JPYyBcit3T_pRtlHdtKSeVJkIyWUTozW2aaut25A2xI,673
-trilogy/core/functions.py,sha256=7Pq9jYSJd45L2pxT7AI-_rXVZmeLnmTPp8d1lA4z4Vk,24440
+trilogy/core/functions.py,sha256=rIkZGzw9hpIkXvuqQ1qPWFJO1W_NPYc6T9t0wTZ55M0,24784
 trilogy/core/graph_models.py,sha256=z17EoO8oky2QOuO6E2aMWoVNKEVJFhLdsQZOhC4fNLU,2079
 trilogy/core/internal.py,sha256=iicDBlC6nM8d7e7jqzf_ZOmpUsW8yrr2AA8AqEiLx-s,1577
 trilogy/core/optimization.py,sha256=xGO8piVsLrpqrx-Aid_Y56_5slSv4eZmlP64hCHRiEc,7957
 trilogy/core/query_processor.py,sha256=Do8YpdPBdsbKtl9n37hobzk8SORMGqH-e_zNNxd-BE4,19456
 trilogy/core/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-trilogy/core/models/author.py,sha256=oRCKWhz-i1fO1LlHWiHE3l1awCHdQ3yx6FKH9n9RxRU,67188
-trilogy/core/models/build.py,sha256=kiq31T8LtUtgmT37m617Q2MlMvQTuAxJzwb6947EiWU,56127
+trilogy/core/models/author.py,sha256=S6riJx8Z4_orbohjK3fcZh5Ei6nCPYAS2FGDsI2njHk,69488
+trilogy/core/models/build.py,sha256=z2QO7l2E2-1hHimmOGsLl42sTnEb2x9o45zkvOoJYpM,56687
 trilogy/core/models/build_environment.py,sha256=8UggvlPU708GZWYPJMc_ou2r7M3TY2g69eqGvz03YX0,5528
 trilogy/core/models/core.py,sha256=yie1uuq62uOQ5fjob9NMJbdvQPrCErXUT7JTCuYRyjI,9697
-trilogy/core/models/datasource.py,sha256=c0tGxyH2WwTmAD047tr69U0a6GNVf-ug26H68yii7DA,9257
-trilogy/core/models/environment.py,sha256=h06y1Dv7naw2GuFFAAyoFZmicG7a7Lu-dRoYPVfrOGo,25967
+trilogy/core/models/datasource.py,sha256=6RjJUd2u4nYmEwFBpJlM9LbHVYDv8iHJxqiBMZqUrwI,9422
+trilogy/core/models/environment.py,sha256=qFZ0_Op6zIhKc5oVS4EVYZ67f29wJhKP_xoEMV4kkuU,25991
 trilogy/core/models/execute.py,sha256=ABylFQgtavjjCfFkEsFdUwfMB4UBQLHjdzQ9E67QlAE,33521
 trilogy/core/optimizations/__init__.py,sha256=EBanqTXEzf1ZEYjAneIWoIcxtMDite5-n2dQ5xcfUtg,356
 trilogy/core/optimizations/base_optimization.py,sha256=gzDOKImoFn36k7XBD3ysEYDnbnb6vdVIztUfFQZsGnM,513
@@ -40,9 +40,9 @@ trilogy/core/processing/node_generators/__init__.py,sha256=o8rOFHPSo-s_59hREwXMW
 trilogy/core/processing/node_generators/basic_node.py,sha256=UVsXMn6jTjm_ofVFt218jAS11s4RV4zD781vP4im-GI,3371
 trilogy/core/processing/node_generators/common.py,sha256=ZsDzThjm_mAtdQpKAg8QIJiPVZ4KuUkKyilj4eOhSDs,9439
 trilogy/core/processing/node_generators/filter_node.py,sha256=rlY7TbgjJlGhahYgdCIJpJbaSREAGVJEsyUIGaA38O0,8271
-trilogy/core/processing/node_generators/group_node.py,sha256=94uoGZWvBKJ1eqjbDHCbZuRqMkux_lfpfkGZgAJTNCY,5876
+trilogy/core/processing/node_generators/group_node.py,sha256=3-TXVnRO9_jqE_e1kWLqbgtBShW8WFtKwQk8oOtOULs,5894
 trilogy/core/processing/node_generators/group_to_node.py,sha256=E5bEjovSx422d_MlAUCDFdY4P2WJVp61BmWwltkhzA8,3095
-trilogy/core/processing/node_generators/multiselect_node.py,sha256=z9FQOxxUvxW31a0TckFfAvnuvU8vP1GyN224RTbXUAk,7114
+trilogy/core/processing/node_generators/multiselect_node.py,sha256=GWV5yLmKTe1yyPhN60RG1Rnrn4ktfn9lYYXi_FVU4UI,7061
 trilogy/core/processing/node_generators/node_merge_node.py,sha256=sv55oynfqgpHEpo1OEtVDri-5fywzPhDlR85qaWikvY,16195
 trilogy/core/processing/node_generators/rowset_node.py,sha256=8yeMWiyi9IFnza7qPn9YYC3WpA53weq3AY5WisIui8Y,6705
 trilogy/core/processing/node_generators/select_merge_node.py,sha256=VHCPMbnKfg7AOfoYa6PKxpNni-j5JEfliNUiltmZhds,19698
@@ -63,7 +63,7 @@ trilogy/core/processing/nodes/union_node.py,sha256=fDFzLAUh5876X6_NM7nkhoMvHEdGJ
 trilogy/core/processing/nodes/unnest_node.py,sha256=oLKMMNMx6PLDPlt2V5neFMFrFWxET8r6XZElAhSNkO0,2181
 trilogy/core/processing/nodes/window_node.py,sha256=STvwheVttxSWVHB-yUQUSo-Pyz7Uk8G1txFDAbWMp-s,1380
 trilogy/core/statements/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-trilogy/core/statements/author.py,sha256=p3gLiPzXAHNNWVh8Xm9xECmywfG-LKDHB9U-Z6GdWCM,14246
+trilogy/core/statements/author.py,sha256=9wKZDwQ-BeaUCMjD9l0ffMMv8zivaYcAg12UhVFi-0Y,14248
 trilogy/core/statements/build.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/core/statements/common.py,sha256=KxEmz2ySySyZ6CTPzn0fJl5NX2KOk1RPyuUSwWhnK1g,759
 trilogy/core/statements/execute.py,sha256=cSlvpHFOqpiZ89pPZ5GDp9Hu6j6uj-5_h21FWm_L-KM,1248
@@ -71,9 +71,10 @@ trilogy/dialect/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/dialect/base.py,sha256=u00kIIl98as1QzcduiiyyoBzxRGVeBxfeO5hWlRCAJU,40222
 trilogy/dialect/bigquery.py,sha256=mKC3zoEU232h9RtIXJjqiZ72lWH8a6S28p6wAZKrAfg,2952
 trilogy/dialect/common.py,sha256=cbTo_vamdp8pj9spSjGSH-bSZpy4FpNJ12k5vMvyT2Y,3942
-trilogy/dialect/config.py,sha256=UiBY2tBbNk9owx-zxP_3lN9lErEUXhXIU_bcXA18AvU,2992
+trilogy/dialect/config.py,sha256=EGYRQIbrkeMuud5Bkds7jSD5dCJR5hEYZUYcy-lYZl4,3308
+trilogy/dialect/dataframe.py,sha256=RUbNgReEa9g3pL6H7fP9lPTrAij5pkqedpZ99D8_5AE,1522
 trilogy/dialect/duckdb.py,sha256=2tH_OetgLJoKf_f4bdeeB0ozGC8f0h_xQ271I8qD-Oo,3690
-trilogy/dialect/enums.py,sha256=JICGp0KQhxNuVDSU36mve2XTZLbPpFRWUDkjvwni4eM,3979
+trilogy/dialect/enums.py,sha256=1KDgds_DC31hGxZzNI_TIggxXF7m9rIjn9KLgNf5WQU,4425
 trilogy/dialect/postgres.py,sha256=VH4EB4myjIeZTHeFU6vK00GxY9c53rCBjg2mLbdaCEE,3254
 trilogy/dialect/presto.py,sha256=bAxaDcLL21fivPg7hmBd3HJmd0yYJdPdwNgNA5ga7DE,3391
 trilogy/dialect/snowflake.py,sha256=wmao9p26jX5yIX5SC8sRAZTXkPGTvq6ixO693QTfhz8,2989
@@ -84,18 +85,18 @@ trilogy/hooks/graph_hook.py,sha256=c-vC-IXoJ_jDmKQjxQyIxyXPOuUcLIURB573gCsAfzQ,2
 trilogy/hooks/query_debugger.py,sha256=1npRjww94sPV5RRBBlLqMJRaFkH9vhEY6o828MeoEcw,5583
 trilogy/metadata/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/parsing/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-trilogy/parsing/common.py,sha256=yAE3x4SyO4PfAb7HhZ_l9sNPYaf_pcM1K8ioEy76SCU,20301
+trilogy/parsing/common.py,sha256=RwO9CdNYy3KxJCjg5Ta_xJwfZHV2PuRErxg66Cs50ws,20490
 trilogy/parsing/config.py,sha256=Z-DaefdKhPDmSXLgg5V4pebhSB0h590vI0_VtHnlukI,111
 trilogy/parsing/exceptions.py,sha256=92E5i2frv5hj9wxObJZsZqj5T6bglvPzvdvco_vW1Zk,38
 trilogy/parsing/helpers.py,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-trilogy/parsing/parse_engine.py,sha256=32_yO_SreTjHxCkMziW2re15ilEZn01OUizVAvN9xHo,54656
+trilogy/parsing/parse_engine.py,sha256=uZ6MYjg6kkTm5HFfOKLGvVvzHiGgH-vY7lV-AIlIBgY,55701
 trilogy/parsing/render.py,sha256=o_XuQWhcwx1lD9eGVqkqZEwkmQK0HdmWWokGBtdeH4I,17837
-trilogy/parsing/trilogy.lark,sha256=EazfEvYPuvkPkNjUnVzFi0uD9baavugbSI8CyfawShk,12573
+trilogy/parsing/trilogy.lark,sha256=wZpqI1louDqm-t-TpmzW749dPA9w2EIAyowyEJIeXAM,12620
 trilogy/scripts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 trilogy/scripts/trilogy.py,sha256=1L0XrH4mVHRt1C9T1HnaDv2_kYEfbWTb5_-cBBke79w,3774
-pytrilogy-0.0.3.6.dist-info/LICENSE.md,sha256=5ZRvtTyCCFwz1THxDTjAu3Lidds9WjPvvzgVwPSYNDo,1042
-pytrilogy-0.0.3.6.dist-info/METADATA,sha256=PBxZLl7AH82ztlgGbJbdLWwj8r3Wo2_JRsXxgh5y1Gc,8983
-pytrilogy-0.0.3.6.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-pytrilogy-0.0.3.6.dist-info/entry_points.txt,sha256=0petKryjvvtEfTlbZC1AuMFumH_WQ9v8A19LvoS6G6c,54
-pytrilogy-0.0.3.6.dist-info/top_level.txt,sha256=cAy__NW_eMAa_yT9UnUNlZLFfxcg6eimUAZ184cdNiE,8
-pytrilogy-0.0.3.6.dist-info/RECORD,,
+pytrilogy-0.0.3.8.dist-info/LICENSE.md,sha256=5ZRvtTyCCFwz1THxDTjAu3Lidds9WjPvvzgVwPSYNDo,1042
+pytrilogy-0.0.3.8.dist-info/METADATA,sha256=DAu0XOCyEgXpZj9-Znl0IbtouGXHELjt2EUZnp3IgEs,8983
+pytrilogy-0.0.3.8.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
+pytrilogy-0.0.3.8.dist-info/entry_points.txt,sha256=0petKryjvvtEfTlbZC1AuMFumH_WQ9v8A19LvoS6G6c,54
+pytrilogy-0.0.3.8.dist-info/top_level.txt,sha256=cAy__NW_eMAa_yT9UnUNlZLFfxcg6eimUAZ184cdNiE,8
+pytrilogy-0.0.3.8.dist-info/RECORD,,

trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
 
-__version__ = "0.0.3.6"
+__version__ = "0.0.3.8"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

trilogy/core/functions.py

@@ -503,32 +503,52 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         arg_count=1,
     ),
     FunctionType.ADD: FunctionConfig(
-        valid_inputs={DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+        valid_inputs={
+            DataType.INTEGER,
+            DataType.FLOAT,
+            DataType.NUMBER,
+            DataType.NUMERIC,
+        },
         output_purpose=Purpose.PROPERTY,
         output_type=DataType.INTEGER,
         arg_count=InfiniteFunctionArgs,
     ),
     FunctionType.SUBTRACT: FunctionConfig(
-        valid_inputs={DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+        valid_inputs={
+            DataType.INTEGER,
+            DataType.FLOAT,
+            DataType.NUMBER,
+            DataType.NUMERIC,
+        },
         output_purpose=Purpose.PROPERTY,
         output_type=DataType.INTEGER,
         arg_count=InfiniteFunctionArgs,
     ),
     FunctionType.MULTIPLY: FunctionConfig(
-        valid_inputs={DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+        valid_inputs={
+            DataType.INTEGER,
+            DataType.FLOAT,
+            DataType.NUMBER,
+            DataType.NUMERIC,
+        },
         output_purpose=Purpose.PROPERTY,
         output_type=DataType.INTEGER,
         arg_count=InfiniteFunctionArgs,
     ),
     FunctionType.DIVIDE: FunctionConfig(
-        valid_inputs={DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+        valid_inputs={
+            DataType.INTEGER,
+            DataType.FLOAT,
+            DataType.NUMBER,
+            DataType.NUMERIC,
+        },
         output_purpose=Purpose.PROPERTY,
         output_type=DataType.INTEGER,
         arg_count=InfiniteFunctionArgs,
     ),
     FunctionType.MOD: FunctionConfig(
         valid_inputs=[
-            {DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+            {DataType.INTEGER, DataType.FLOAT, DataType.NUMBER, DataType.NUMERIC},
             {DataType.INTEGER},
         ],
         output_purpose=Purpose.PROPERTY,
@@ -537,7 +557,7 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
     ),
     FunctionType.ROUND: FunctionConfig(
         valid_inputs=[
-            {DataType.INTEGER, DataType.FLOAT, DataType.NUMBER},
+            {DataType.INTEGER, DataType.FLOAT, DataType.NUMBER, DataType.NUMERIC},
             {DataType.INTEGER},
         ],
         output_purpose=Purpose.PROPERTY,

trilogy/core/models/author.py

@@ -75,6 +75,9 @@ class Mergeable(ABC):
     def with_merge(self, source: Concept, target: Concept, modifiers: List[Modifier]):
         raise NotImplementedError
 
+    def with_reference_replacement(self, source: str, target: Expr):
+        raise NotImplementedError(type(self))
+
 
 class ConceptArgs(ABC):
     @property
@@ -152,6 +155,11 @@ class ConceptRef(Addressable, Namespaced, DataTyped, Mergeable, BaseModel):
             metadata=self.metadata,
         )
 
+    def with_reference_replacement(self, source: str, target: Expr):
+        if self.address == source:
+            return target
+        return self
+
 
 class UndefinedConcept(ConceptRef):
     pass
@@ -409,7 +417,7 @@ class Grain(Namespaced, BaseModel):
     ) -> Grain:
         from trilogy.parsing.common import concepts_to_grain_concepts
 
-        return Grain.model_construct(
+        x = Grain.model_construct(
             components={
                 c.address
                 for c in concepts_to_grain_concepts(concepts, environment=environment)
@@ -417,6 +425,8 @@ class Grain(Namespaced, BaseModel):
             where_clause=where_clause,
         )
 
+        return x
+
     def with_namespace(self, namespace: str) -> "Grain":
         return Grain.model_construct(
             components={address_with_namespace(c, namespace) for c in self.components},
@@ -662,6 +672,21 @@ class Comparison(ConceptArgs, Mergeable, Namespaced, BaseModel):
             operator=self.operator,
         )
 
+    def with_reference_replacement(self, source, target):
+        return self.__class__.model_construct(
+            left=(
+                self.left.with_reference_replacement(source, target)
+                if isinstance(self.left, Mergeable)
+                else self.left
+            ),
+            right=(
+                self.right.with_reference_replacement(source, target)
+                if isinstance(self.right, Mergeable)
+                else self.right
+            ),
+            operator=self.operator,
+        )
+
     def with_namespace(self, namespace: str):
         return self.__class__.model_construct(
             left=(
@@ -1233,19 +1258,9 @@ class OrderItem(Mergeable, ConceptArgs, Namespaced, BaseModel):
             order=self.order,
         )
 
-    @property
-    def output(self):
-        return self.expr.output
-
     @property
     def concept_arguments(self) -> Sequence[ConceptRef]:
-        base: List[ConceptRef] = []
-        x = self.expr
-        if isinstance(x, ConceptRef):
-            base += [x]
-        elif isinstance(x, ConceptArgs):
-            base += x.concept_arguments
-        return base
+        return get_concept_arguments(self.expr)
 
     @property
     def row_arguments(self) -> Sequence[ConceptRef]:
@@ -1316,31 +1331,19 @@ class WindowItem(DataTyped, ConceptArgs, Mergeable, Namespaced, BaseModel):
 
     @property
     def concept_arguments(self) -> List[ConceptRef]:
-        return self.arguments
-
-    @property
-    def arguments(self) -> List[ConceptRef]:
         output = [self.content]
         for order in self.order_by:
-            output += [order.output]
+            output += get_concept_arguments(order)
         for item in self.over:
-            output += [item]
+            output += get_concept_arguments(item)
         return output
 
-    @property
-    def output(self) -> ConceptRef:
-        return self.content
-
     @property
     def output_datatype(self):
         if self.type in (WindowType.RANK, WindowType.ROW_NUMBER):
             return DataType.INTEGER
         return self.content.output_datatype
 
-    @property
-    def output_purpose(self):
-        return Purpose.PROPERTY
-
 
 def get_basic_type(
     type: DataType | ListType | StructType | MapType | NumericType,
@@ -1407,12 +1410,28 @@ class CaseWhen(Namespaced, ConceptArgs, Mergeable, BaseModel):
             ),
         )
 
+    def with_reference_replacement(self, source, target):
+        return CaseWhen.model_construct(
+            comparison=self.comparison.with_reference_replacement(source, target),
+            expr=(
+                self.expr.with_reference_replacement(source, target)
+                if isinstance(self.expr, Mergeable)
+                else self.expr
+            ),
+        )
+
 
 class CaseElse(Namespaced, ConceptArgs, Mergeable, BaseModel):
     expr: "Expr"
     # this ensures that it's easily differentiable from CaseWhen
     discriminant: ComparisonOperator = ComparisonOperator.ELSE
 
+    def __str__(self):
+        return self.__repr__()
+
+    def __repr__(self):
+        return f"ELSE {str(self.expr)}"
+
     @field_validator("expr", mode="before")
     def enforce_expr(cls, v):
         if isinstance(v, Concept):
@@ -1435,6 +1454,19 @@ class CaseElse(Namespaced, ConceptArgs, Mergeable, BaseModel):
             ),
         )
 
+    def with_reference_replacement(self, source, target):
+        return CaseElse.model_construct(
+            discriminant=self.discriminant,
+            expr=(
+                self.expr.with_reference_replacement(
+                    source,
+                    target,
+                )
+                if isinstance(self.expr, Mergeable)
+                else self.expr
+            ),
+        )
+
     def with_namespace(self, namespace: str) -> CaseElse:
         return CaseElse.model_construct(
             discriminant=self.discriminant,
@@ -1510,8 +1542,6 @@ class Function(DataTyped, ConceptArgs, Mergeable, Namespaced, BaseModel):
 
     def __init__(self, **kwargs):
         super().__init__(**kwargs)
-        if "datatype" in str(self):
-            raise SyntaxError(str(self))
 
     def __repr__(self):
         return f'{self.operator.value}({",".join([str(a) for a in self.arguments])})'
@@ -1597,6 +1627,29 @@ class Function(DataTyped, ConceptArgs, Mergeable, Namespaced, BaseModel):
                     )
         return v
 
+    def with_reference_replacement(self, source: str, target: Expr):
+        return Function.model_construct(
+            operator=self.operator,
+            arguments=[
+                (
+                    c.with_reference_replacement(
+                        source,
+                        target,
+                    )
+                    if isinstance(
+                        c,
+                        Mergeable,
+                    )
+                    else c
+                )
+                for c in self.arguments
+            ],
+            output_datatype=self.output_datatype,
+            output_purpose=self.output_purpose,
+            valid_inputs=self.valid_inputs,
+            arg_count=self.arg_count,
+        )
+
     def with_namespace(self, namespace: str) -> "Function":
         return Function.model_construct(
             operator=self.operator,
@@ -1658,10 +1711,13 @@ class Function(DataTyped, ConceptArgs, Mergeable, Namespaced, BaseModel):
         return base_grain
 
 
-class AggregateWrapper(Mergeable, ConceptArgs, Namespaced, BaseModel):
+class AggregateWrapper(Mergeable, DataTyped, ConceptArgs, Namespaced, BaseModel):
     function: Function
     by: List[ConceptRef] = Field(default_factory=list)
 
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+
     @field_validator("by", mode="before")
     @classmethod
     def enforce_concept_ref(cls, v):
@@ -1693,10 +1749,6 @@ class AggregateWrapper(Mergeable, ConceptArgs, Namespaced, BaseModel):
     def output_purpose(self):
         return self.function.output_purpose
 
-    @property
-    def arguments(self):
-        return self.function.arguments
-
     def with_merge(self, source: Concept, target: Concept, modifiers: List[Modifier]):
         return AggregateWrapper.model_construct(
             function=self.function.with_merge(source, target, modifiers=modifiers),
@@ -1707,6 +1759,16 @@ class AggregateWrapper(Mergeable, ConceptArgs, Namespaced, BaseModel):
             ),
         )
 
+    def with_reference_replacement(self, source, target):
+        return AggregateWrapper.model_construct(
+            function=self.function.with_reference_replacement(source, target),
+            by=(
+                [c.with_reference_replacement(source, target) for c in self.by]
+                if self.by
+                else []
+            ),
+        )
+
     def with_namespace(self, namespace: str) -> "AggregateWrapper":
         return AggregateWrapper.model_construct(
             function=self.function.with_namespace(namespace),
@@ -1796,11 +1858,6 @@ class RowsetItem(Mergeable, ConceptArgs, Namespaced, BaseModel):
             rowset=self.rowset.with_namespace(namespace),
         )
 
-    @property
-    def arguments(self) -> List[ConceptRef]:
-        output = [self.content]
-        return output
-
     @property
     def output(self) -> ConceptRef:
         return self.content
@@ -1831,7 +1888,10 @@ class OrderBy(Mergeable, Namespaced, BaseModel):
 
     @property
     def concept_arguments(self):
-        return [x.expr for x in self.items]
+        base = []
+        for x in self.items:
+            base += x.concept_arguments
+        return base
 
 
 class AlignClause(Namespaced, BaseModel):
@@ -2088,6 +2148,11 @@ class Comment(BaseModel):
     text: str
 
 
+class ArgBinding(BaseModel):
+    name: str
+    default: Expr | None = None
+
+
 Expr = (
     MagicConstants
     | bool

trilogy/core/models/build.py

@@ -1639,9 +1639,26 @@ class Factory:
 
     @build.register
     def _(self, base: Comparison) -> BuildComparison:
+        from trilogy.parsing.common import arbitrary_to_concept
+
+        left = base.left
+        if isinstance(left, AggregateWrapper):
+            left_c = arbitrary_to_concept(
+                left,
+                environment=self.environment,
+            )
+            left = left_c  # type: ignore
+        right = base.right
+        if isinstance(right, AggregateWrapper):
+            right_c = arbitrary_to_concept(
+                right,
+                environment=self.environment,
+            )
+
+            right = right_c  # type: ignore
         return BuildComparison.model_construct(
-            left=(self.build(base.left)),
-            right=(self.build(base.right)),
+            left=(self.build(left)),
+            right=(self.build(right)),
             operator=base.operator,
         )
 
@@ -1660,7 +1677,7 @@ class Factory:
         )
 
     @build.register
-    def _(self, base: RowsetItem):
+    def _(self, base: RowsetItem) -> BuildRowsetItem:
 
         factory = Factory(
             environment=self.environment,

trilogy/core/models/datasource.py

@@ -117,6 +117,12 @@ class Datasource(HasUUID, Namespaced, BaseModel):
     where: Optional[WhereClause] = None
     non_partial_for: Optional[WhereClause] = None
 
+    @property
+    def safe_address(self) -> str:
+        if isinstance(self.address, Address):
+            return self.address.location
+        return self.address
+
     def __eq__(self, other):
         if not isinstance(other, Datasource):
             return False

trilogy/core/models/environment.py

@@ -9,6 +9,7 @@ from typing import (
     TYPE_CHECKING,
     Annotated,
     Any,
+    Callable,
     Dict,
     ItemsView,
     List,
@@ -189,7 +190,7 @@ class Environment(BaseModel):
     datasources: Annotated[
         EnvironmentDatasourceDict, PlainValidator(validate_datasources)
     ] = Field(default_factory=EnvironmentDatasourceDict)
-    functions: Dict[str, Function] = Field(default_factory=dict)
+    functions: Dict[str, Callable[..., Any]] = Field(default_factory=dict)
     data_types: Dict[str, DataType] = Field(default_factory=dict)
     named_statements: Dict[str, SelectLineage] = Field(default_factory=dict)
     imports: Dict[str, list[Import]] = Field(

trilogy/core/processing/node_generators/group_node.py

@@ -37,7 +37,7 @@ def gen_group_node(
         resolve_function_parent_concepts(concept, environment=environment), "address"
     )
     logger.info(
-        f"{padding(depth)}{LOGGER_PREFIX} parent concepts for {concept} are {[x.address for x in parent_concepts]} from group grain {concept.grain}"
+        f"{padding(depth)}{LOGGER_PREFIX} parent concepts for {concept} {concept.lineage} are {[x.address for x in parent_concepts]} from group grain {concept.grain}"
     )
 
     # if the aggregation has a grain, we need to ensure these are the ONLY optional in the output of the select

trilogy/core/processing/node_generators/multiselect_node.py

@@ -76,7 +76,7 @@ def gen_multiselect_node(
     for select in lineage.selects:
 
         snode: StrategyNode = get_query_node(history.base_environment, select)
-        # raise SyntaxError(select.output_components)
+
         logger.info(
             f"{padding(depth)}{LOGGER_PREFIX} Fetched parent node with outputs {select.output_components}"
         )

trilogy/core/statements/author.py

@@ -127,7 +127,9 @@ class SelectStatement(HasUUID, SelectTypeMixin, BaseModel):
             order_by=order_by,
             meta=meta or Metadata(),
         )
+
         output.grain = output.calculate_grain(environment)
+
         for x in selection:
 
             if x.is_undefined and environment.concepts.fail_on_missing:

trilogy/dialect/config.py

@@ -1,3 +1,12 @@
+from typing import TYPE_CHECKING, Any
+
+if TYPE_CHECKING:
+    try:
+        from pandas import DataFrame
+    except ImportError:
+        DataFrame = Any
+
+
 class DialectConfig:
     def __init__(self):
         pass
@@ -104,3 +113,9 @@ class TrinoConfig(PrestoConfig):
         if self.schema:
             return f"trino://{self.username}:{self.password}@{self.host}:{self.port}/{self.catalog}/{self.schema}"
         return f"trino://{self.username}:{self.password}@{self.host}:{self.port}/{self.catalog}"
+
+
+class DataFrameConfig(DuckDBConfig):
+    def __init__(self, dataframes: dict[str, "DataFrame"]):
+        super().__init__()
+        self.dataframes = dataframes

trilogy/dialect/dataframe.py

@@ -0,0 +1,47 @@
+from typing import TYPE_CHECKING, Any
+
+from sqlalchemy import text
+
+from trilogy.core.models.environment import Environment
+from trilogy.dialect.duckdb import DuckDBDialect
+from trilogy.engine import ExecutionEngine
+
+if TYPE_CHECKING:
+    try:
+        from pandas import DataFrame
+    except ImportError:
+        DataFrame = Any
+
+
+class DataframeDialect(DuckDBDialect):
+    pass
+
+
+class DataframeConnectionWrapper(ExecutionEngine):
+    def __init__(self, engine: ExecutionEngine, dataframes: dict[str, "DataFrame"]):
+        self.engine = engine
+        self.dataframes = dataframes
+        self.connection = None
+
+    def setup(self, env: Environment, connection):
+        self._register_dataframes(env, connection)
+
+    def _register_dataframes(self, env: Environment, connection):
+        for ds in env.datasources.values():
+            if ds.safe_address in self.dataframes:
+                connection.execute(
+                    text("register(:name, :df)"),
+                    {"name": ds.safe_address, "df": self.dataframes[ds.safe_address]},
+                )
+            else:
+                raise ValueError(
+                    f"Dataframe {ds.safe_address} not found in dataframes on connection config, have {self.dataframes.keys()}"
+                )
+        pass
+
+    def add_dataframe(self, name: str, df: "DataFrame", connection, env: Environment):
+        self.dataframes[name] = df
+        self._register_dataframes(env, connection)
+
+    def connect(self) -> Any:
+        return self.engine.connect()

trilogy/dialect/enums.py

@@ -16,7 +16,7 @@ def default_factory(conf: DialectConfig, config_type):
 
     if not isinstance(conf, config_type):
         raise TypeError(
-            f"Invalid dialect configuration for type {type(config_type).__name__}"
+            f"Invalid dialect configuration for type {type(config_type).__name__}, is {type(conf)}"
         )
     if conf.connect_args:
         return create_engine(
@@ -33,6 +33,7 @@ class Dialects(Enum):
     TRINO = "trino"
     POSTGRES = "postgres"
     SNOWFLAKE = "snowflake"
+    DATAFRAME = "dataframe"
 
     @classmethod
     def _missing_(cls, value):
@@ -88,6 +89,16 @@ class Dialects(Enum):
             from trilogy.dialect.config import TrinoConfig
 
             return _engine_factory(conf, TrinoConfig)
+        elif self == Dialects.DATAFRAME:
+            from trilogy.dialect.config import DataFrameConfig
+            from trilogy.dialect.dataframe import DataframeConnectionWrapper
+
+            if not conf:
+                conf = DataFrameConfig(dataframes={})
+
+            base = _engine_factory(conf, DataFrameConfig)
+
+            return DataframeConnectionWrapper(base, dataframes=conf.dataframes)
         else:
             raise ValueError(
                 f"Unsupported dialect {self} for default engine creation; create one explicitly."

trilogy/engine.py

@@ -1,7 +1,9 @@
-from typing import Protocol
+from typing import Any, Protocol
 
 from sqlalchemy.engine import Connection, CursorResult, Engine
 
+from trilogy.core.models.environment import Environment
+
 
 class EngineResult(Protocol):
     pass
@@ -13,7 +15,7 @@ class EngineResult(Protocol):
 class EngineConnection(Protocol):
     pass
 
-    def execute(self, statement: str) -> EngineResult:
+    def execute(self, statement: str, parameters: Any | None = None) -> EngineResult:
         pass
 
 
@@ -23,6 +25,9 @@ class ExecutionEngine(Protocol):
     def connect(self) -> EngineConnection:
         pass
 
+    def setup(self, env: Environment, connection):
+        pass
+
 
 ### Begin default SQLAlchemy implementation
 class SqlAlchemyResult(EngineResult):
@@ -37,8 +42,10 @@ class SqlAlchemyConnection(EngineConnection):
     def __init__(self, connection: Connection):
         self.connection = connection
 
-    def execute(self, statement: str) -> SqlAlchemyResult:
-        return SqlAlchemyResult(self.connection.execute(statement))
+    def execute(
+        self, statement: str, parameters: Any | None = None
+    ) -> SqlAlchemyResult:
+        return SqlAlchemyResult(self.connection.execute(statement, parameters))
 
 
 class SqlAlchemyEngine(ExecutionEngine):

trilogy/executor.py

@@ -4,7 +4,7 @@ from pathlib import Path
 from typing import Any, Generator, List, Optional, Protocol
 
 from sqlalchemy import text
-from sqlalchemy.engine import CursorResult, Engine
+from sqlalchemy.engine import CursorResult
 
 from trilogy.constants import logger
 from trilogy.core.enums import FunctionType, Granularity, IOType
@@ -33,6 +33,7 @@ from trilogy.core.statements.execute import (
 )
 from trilogy.dialect.base import BaseDialect
 from trilogy.dialect.enums import Dialects
+from trilogy.engine import ExecutionEngine
 from trilogy.hooks.base_hook import BaseHook
 from trilogy.parser import parse_text
 
@@ -71,7 +72,7 @@ class Executor(object):
     def __init__(
         self,
         dialect: Dialects,
-        engine: Engine,
+        engine: ExecutionEngine,
         environment: Optional[Environment] = None,
         hooks: List[BaseHook] | None = None,
     ):
@@ -109,9 +110,16 @@ class Executor(object):
             from trilogy.dialect.snowflake import SnowflakeDialect
 
             self.generator = SnowflakeDialect()
+        elif self.dialect == Dialects.DATAFRAME:
+            from trilogy.dialect.dataframe import DataframeDialect
+
+            self.generator = DataframeDialect()
         else:
             raise ValueError(f"Unsupported dialect {self.dialect}")
         self.connection = self.engine.connect()
+        # TODO: make generic
+        if self.dialect == Dialects.DATAFRAME:
+            self.engine.setup(self.environment, self.connection)
 
     def execute_statement(self, statement) -> Optional[CursorResult]:
         if not isinstance(

trilogy/parsing/common.py

@@ -46,6 +46,62 @@ from trilogy.core.statements.author import RowsetDerivationStatement, SelectStat
 from trilogy.utility import string_to_hash, unique
 
 
+def process_function_arg(
+    arg,
+    meta: Meta | None,
+    environment: Environment,
+):
+    # if a function has an anonymous function argument
+    # create an implicit concept
+    if isinstance(arg, Parenthetical):
+        processed = process_function_args([arg.content], meta, environment)
+        return Function(
+            operator=FunctionType.PARENTHETICAL,
+            arguments=processed,
+            output_datatype=arg_to_datatype(processed[0]),
+            output_purpose=function_args_to_output_purpose(processed),
+        )
+    elif isinstance(arg, Function):
+        # if it's not an aggregate function, we can skip the virtual concepts
+        # to simplify anonymous function handling
+        if (
+            arg.operator not in FunctionClass.AGGREGATE_FUNCTIONS.value
+            and arg.operator != FunctionType.UNNEST
+        ):
+            return arg
+        id_hash = string_to_hash(str(arg))
+        name = f"{VIRTUAL_CONCEPT_PREFIX}_{arg.operator.value}_{id_hash}"
+        if f"{environment.namespace}.{name}" in environment.concepts:
+            return environment.concepts[f"{environment.namespace}.{name}"]
+        concept = function_to_concept(
+            arg,
+            name=name,
+            environment=environment,
+        )
+        # to satisfy mypy, concept will always have metadata
+        if concept.metadata and meta:
+            concept.metadata.line_number = meta.line
+        environment.add_concept(concept, meta=meta)
+        return concept
+    elif isinstance(
+        arg, (FilterItem, WindowItem, AggregateWrapper, ListWrapper, MapWrapper)
+    ):
+        id_hash = string_to_hash(str(arg))
+        name = f"{VIRTUAL_CONCEPT_PREFIX}_{id_hash}"
+        if f"{environment.namespace}.{name}" in environment.concepts:
+            return environment.concepts[f"{environment.namespace}.{name}"]
+        concept = arbitrary_to_concept(
+            arg,
+            name=name,
+            environment=environment,
+        )
+        if concept.metadata and meta:
+            concept.metadata.line_number = meta.line
+        environment.add_concept(concept, meta=meta)
+        return concept
+    return arg
+
+
 def process_function_args(
     args,
     meta: Meta | None,
@@ -53,54 +109,7 @@ def process_function_args(
 ) -> List[Concept | Function | str | int | float | date | datetime]:
     final: List[Concept | Function | str | int | float | date | datetime] = []
     for arg in args:
-        # if a function has an anonymous function argument
-        # create an implicit concept
-        if isinstance(arg, Parenthetical):
-            processed = process_function_args([arg.content], meta, environment)
-            final.append(
-                Function(
-                    operator=FunctionType.PARENTHETICAL,
-                    arguments=processed,
-                    output_datatype=arg_to_datatype(processed[0]),
-                    output_purpose=function_args_to_output_purpose(processed),
-                )
-            )
-        elif isinstance(arg, Function):
-            # if it's not an aggregate function, we can skip the virtual concepts
-            # to simplify anonymous function handling
-            if (
-                arg.operator not in FunctionClass.AGGREGATE_FUNCTIONS.value
-                and arg.operator != FunctionType.UNNEST
-            ):
-                final.append(arg)
-                continue
-            id_hash = string_to_hash(str(arg))
-            concept = function_to_concept(
-                arg,
-                name=f"{VIRTUAL_CONCEPT_PREFIX}_{arg.operator.value}_{id_hash}",
-                environment=environment,
-            )
-            # to satisfy mypy, concept will always have metadata
-            if concept.metadata and meta:
-                concept.metadata.line_number = meta.line
-            environment.add_concept(concept, meta=meta)
-            final.append(concept)
-        elif isinstance(
-            arg, (FilterItem, WindowItem, AggregateWrapper, ListWrapper, MapWrapper)
-        ):
-            id_hash = string_to_hash(str(arg))
-            concept = arbitrary_to_concept(
-                arg,
-                name=f"{VIRTUAL_CONCEPT_PREFIX}_{id_hash}",
-                environment=environment,
-            )
-            if concept.metadata and meta:
-                concept.metadata.line_number = meta.line
-            environment.add_concept(concept, meta=meta)
-            final.append(concept)
-
-        else:
-            final.append(arg)
+        final.append(process_function_arg(arg, meta, environment))
     return final
 
 

trilogy/parsing/parse_engine.py

@@ -4,7 +4,7 @@ from enum import Enum
 from os.path import dirname, join
 from pathlib import Path
 from re import IGNORECASE
-from typing import List, Optional, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 from lark import Lark, ParseTree, Transformer, Tree, v_args
 from lark.exceptions import (
@@ -49,6 +49,7 @@ from trilogy.core.models.author import (
     AggregateWrapper,
     AlignClause,
     AlignItem,
+    ArgBinding,
     CaseElse,
     CaseWhen,
     Comment,
@@ -61,6 +62,7 @@ from trilogy.core.models.author import (
     Function,
     Grain,
     HavingClause,
+    Mergeable,
     Metadata,
     OrderBy,
     OrderItem,
@@ -1075,27 +1077,45 @@ class ParseToObjects(Transformer):
         return HavingClause(conditional=root)
 
     @v_args(meta=True)
-    def function_binding_list(self, meta: Meta, args) -> Concept:
+    def function_binding_list(self, meta: Meta, args) -> list[ArgBinding]:
         return args
 
     @v_args(meta=True)
-    def function_binding_item(self, meta: Meta, args) -> Concept:
-        return args
+    def function_binding_item(self, meta: Meta, args) -> ArgBinding:
+        if len(args) == 2:
+            return ArgBinding(name=args[0], default=args[1])
+        return ArgBinding(name=args[0], default=None)
 
     @v_args(meta=True)
-    def raw_function(self, meta: Meta, args) -> Function:
+    def raw_function(self, meta: Meta, args) -> Callable[[list[Expr]], Expr]:
         identity = args[0]
-        fargs = args[1]
+        function_arguments: list[ArgBinding] = args[1]
         output = args[2]
-        item = Function(
-            operator=FunctionType.SUM,
-            arguments=[x[1] for x in fargs],
-            output_datatype=output,
-            output_purpose=Purpose.PROPERTY,
-            arg_count=len(fargs) + 1,
-        )
-        self.environment.functions[identity] = item
-        return item
+
+        def function_factory(*creation_args: list[Expr]):
+            nout = output.copy(deep=True)
+            creation_arg_list: list[Expr] = list(creation_args)
+            if len(creation_args) < len(function_arguments):
+                for binding in function_arguments[len(creation_arg_list) :]:
+                    if binding.default is None:
+                        raise ValueError(f"Missing argument {binding.name}")
+                    creation_arg_list.append(binding.default)
+            if isinstance(nout, Mergeable):
+                for idx, x in enumerate(creation_arg_list):
+                    # these will always be local namespace
+                    nout = nout.with_reference_replacement(
+                        f"{DEFAULT_NAMESPACE}.{function_arguments[idx].name}", x
+                    )
+            return nout
+
+        self.environment.functions[identity] = function_factory
+        return function_factory
+
+    def custom_function(self, args):
+        name = args[0]
+        args = args[1:]
+        remapped = self.environment.functions[name](*args)
+        return remapped
 
     @v_args(meta=True)
     def function(self, meta: Meta, args) -> Function:
@@ -1141,24 +1161,24 @@ class ParseToObjects(Transformer):
     def comparison(self, args) -> Comparison:
         if args[1] == ComparisonOperator.IN:
             raise SyntaxError
-        if isinstance(args[0], AggregateWrapper):
-            left_c = arbitrary_to_concept(
-                args[0],
-                environment=self.environment,
-            )
-            self.environment.add_concept(left_c)
-            left = left_c.reference
-        else:
-            left = args[0]
-        if isinstance(args[2], AggregateWrapper):
-            right_c = arbitrary_to_concept(
-                args[2],
-                environment=self.environment,
-            )
-            self.environment.add_concept(right_c)
-            right = right_c.reference
-        else:
-            right = args[2]
+        # if isinstance(args[0], AggregateWrapper):
+        #     left_c = arbitrary_to_concept(
+        #         args[0],
+        #         environment=self.environment,
+        #     )
+        #     self.environment.add_concept(left_c)
+        #     left = left_c.reference
+        # else:
+        left = args[0]
+        # if isinstance(args[2], AggregateWrapper):
+        #     right_c = arbitrary_to_concept(
+        #         args[2],
+        #         environment=self.environment,
+        #     )
+        #     self.environment.add_concept(right_c)
+        #     right = right_c.reference
+        # else:
+        right = args[2]
         return Comparison(left=left, right=right, operator=args[1])
 
     def between_comparison(self, args) -> Conditional:

trilogy/parsing/trilogy.lark

@@ -89,9 +89,9 @@
 
     // FUNCTION blocks
     function: raw_function
-    function_binding_item: IDENTIFIER ":" data_type
-    function_binding_list: function_binding_item ("," function_binding_item )* ","?
-    raw_function: "bind" "sql"  IDENTIFIER  "("  function_binding_list ")" "->" data_type "as"i MULTILINE_STRING
+    function_binding_item: IDENTIFIER ("=" literal)?
+    function_binding_list: (function_binding_item ",")* function_binding_item ","?
+    raw_function: "def"  IDENTIFIER  "("  function_binding_list ")" "->" expr
     
     // user_id where state = Mexico
     _filter_alt: IDENTIFIER "?" conditional
@@ -178,7 +178,7 @@
     map_key_access: expr "[" string_lit "]"
     attr_access: expr  "." string_lit 
 
-    expr:  _constant_functions | window_item | filter_item | subselect_comparison | between_comparison |  fgroup |  aggregate_functions | unnest | union | _static_functions |  literal |  concept_lit  | index_access |  map_key_access | attr_access | parenthetical | expr_tuple | comparison |  alt_like 
+    expr:  _constant_functions | window_item | filter_item | subselect_comparison | between_comparison |  fgroup |  aggregate_functions | unnest | union | _static_functions |  literal |  concept_lit  | index_access |  map_key_access | attr_access | parenthetical | expr_tuple | comparison |  alt_like | custom_function 
     
     // functions
     
@@ -297,6 +297,8 @@
     
     _static_functions:  _string_functions | _math_functions | _generic_functions |  _date_functions
 
+    custom_function:  "@" IDENTIFIER  "(" (expr ",")* expr ")"
+
     // base language constructs
     concept_lit: IDENTIFIER
     IDENTIFIER: /[a-zA-Z\_][a-zA-Z0-9\_\.]*/