pytrilogy 0.0.3.69__tar.gz → 0.0.3.70__tar.gz

{pytrilogy-0.0.3.69/pytrilogy.egg-info → pytrilogy-0.0.3.70}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.69
+Version: 0.0.3.70
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70/pytrilogy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytrilogy
-Version: 0.0.3.69
+Version: 0.0.3.70
 Summary: Declarative, typed query language that compiles to SQL.
 Home-page: 
 Author: 

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70}/trilogy/__init__.py

@@ -4,6 +4,6 @@ from trilogy.dialect.enums import Dialects
 from trilogy.executor import Executor
 from trilogy.parser import parse
 
-__version__ = "0.0.3.69"
+__version__ = "0.0.3.70"
 
 __all__ = ["parse", "Executor", "Dialects", "Environment", "CONFIG"]

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70}/trilogy/core/functions.py

@@ -370,10 +370,10 @@ FUNCTION_REGISTRY: dict[FunctionType, FunctionConfig] = {
         arg_count=2,
     ),
     FunctionType.REGEXP_EXTRACT: FunctionConfig(
-        valid_inputs={DataType.STRING},
+        valid_inputs=[{DataType.STRING}, {DataType.STRING}, {DataType.INTEGER}],
         output_purpose=Purpose.PROPERTY,
         output_type=DataType.STRING,
-        arg_count=2,
+        arg_count=3,
     ),
     FunctionType.REGEXP_REPLACE: FunctionConfig(
         valid_inputs={DataType.STRING},

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70}/trilogy/dialect/duckdb.py

@@ -1,3 +1,4 @@
+import re
 from typing import Any, Callable, Mapping
 
 from jinja2 import Template
@@ -8,6 +9,19 @@ from trilogy.dialect.base import BaseDialect
 
 WINDOW_FUNCTION_MAP: Mapping[WindowType, Callable[[Any, Any, Any], str]] = {}
 
+SENTINAL_AUTO_CAPTURE_GROUP_VALUE = "-1"
+
+
+def generate_regex_extract(x: list[str]) -> str:
+    if str(x[2]) == SENTINAL_AUTO_CAPTURE_GROUP_VALUE:
+        regex = re.compile(x[1])
+        if regex.groups == 0:
+            search = 0
+        else:
+            search = 1
+        return f"REGEXP_EXTRACT({x[0]},{x[1]},{search})"
+    return f"REGEXP_EXTRACT({x[0]},{x[1]},{x[2]})"
+
 
 FUNCTION_MAP = {
     FunctionType.COUNT: lambda args: f"count({args[0]})",
@@ -37,6 +51,9 @@ FUNCTION_MAP = {
     FunctionType.DATETIME_LITERAL: lambda x: f"datetime '{x}'",
     # string
     FunctionType.CONTAINS: lambda x: f"CONTAINS(LOWER({x[0]}), LOWER({x[1]}))",
+    # regexp
+    FunctionType.REGEXP_CONTAINS: lambda x: f"REGEXP_MATCHES({x[0]},{x[1]})",
+    FunctionType.REGEXP_EXTRACT: lambda x: generate_regex_extract(x),
 }
 
 # if an aggregate function is called on a source that is at the same grain as the aggregate

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70}/trilogy/parsing/parse_engine.py

@@ -1802,6 +1802,9 @@ class ParseToObjects(Transformer):
 
     @v_args(meta=True)
     def fregexp_extract(self, meta, args):
+        if len(args) == 2:
+            # this is a magic value to represent the default behavior
+            args.append(-1)
         return self.function_factory.create_function(
             args, FunctionType.REGEXP_EXTRACT, meta
         )

{pytrilogy-0.0.3.69 → pytrilogy-0.0.3.70}/trilogy/parsing/trilogy.lark

@@ -279,7 +279,7 @@
     _SUBSTRING.1: "substring("i
     fsubstring: _SUBSTRING expr "," expr "," expr ")"
     _REGEXP_EXTRACT.1: "regexp_extract("
-    fregexp_extract: _REGEXP_EXTRACT expr "," expr ")"
+    fregexp_extract: _REGEXP_EXTRACT expr "," expr ("," int_lit)? ")"
     _REGEXP_CONTAINS.1: "regexp_contains("
     fregexp_contains: _REGEXP_CONTAINS expr "," expr  ")"
     _REGEXP_REPLACE.1: "regexp_replace("