Adding upstream version 25.29.0.

Signed-off-by: Daniel Baumann <daniel@debian.org>

sqlglot/expressions.py

@@ -4312,6 +4312,7 @@ class DataType(Expression):
         DECIMAL32 = auto()
         DECIMAL64 = auto()
         DECIMAL128 = auto()
+        DECIMAL256 = auto()
         DOUBLE = auto()
         ENUM = auto()
         ENUM8 = auto()
@@ -4320,6 +4321,12 @@ class DataType(Expression):
         FLOAT = auto()
         GEOGRAPHY = auto()
         GEOMETRY = auto()
+        POINT = auto()
+        RING = auto()
+        LINESTRING = auto()
+        MULTILINESTRING = auto()
+        POLYGON = auto()
+        MULTIPOLYGON = auto()
         HLLSKETCH = auto()
         HSTORE = auto()
         IMAGE = auto()
@@ -4467,6 +4474,7 @@ class DataType(Expression):
         Type.DECIMAL32,
         Type.DECIMAL64,
         Type.DECIMAL128,
+        Type.DECIMAL256,
         Type.MONEY,
         Type.SMALLMONEY,
         Type.UDECIMAL,
@@ -4597,10 +4605,6 @@ class Any(SubqueryPredicate):
     pass
 
 
-class Exists(SubqueryPredicate):
-    pass
-
-
 # Commands to interact with the databases or engines. For most of the command
 # expressions we parse whatever comes after the command's name as a string.
 class Command(Expression):
@@ -5276,6 +5280,11 @@ class ArrayToString(Func):
     _sql_names = ["ARRAY_TO_STRING", "ARRAY_JOIN"]
 
 
+# https://cloud.google.com/bigquery/docs/reference/standard-sql/timestamp_functions#string
+class String(Func):
+    arg_types = {"this": True, "zone": False}
+
+
 class StringToArray(Func):
     arg_types = {"this": True, "expression": True, "null": False}
     _sql_names = ["STRING_TO_ARRAY", "SPLIT_BY_STRING"]
@@ -5494,11 +5503,17 @@ class DateTrunc(Func):
     arg_types = {"unit": True, "this": True, "zone": False}
 
     def __init__(self, **args):
+        # Across most dialects it's safe to unabbreviate the unit (e.g. 'Q' -> 'QUARTER') except Oracle
+        # https://docs.oracle.com/en/database/oracle/oracle-database/21/sqlrf/ROUND-and-TRUNC-Date-Functions.html
+        unabbreviate = args.pop("unabbreviate", True)
+
         unit = args.get("unit")
         if isinstance(unit, TimeUnit.VAR_LIKE):
-            args["unit"] = Literal.string(
-                (TimeUnit.UNABBREVIATED_UNIT_NAME.get(unit.name) or unit.name).upper()
-            )
+            unit_name = unit.name.upper()
+            if unabbreviate and unit_name in TimeUnit.UNABBREVIATED_UNIT_NAME:
+                unit_name = TimeUnit.UNABBREVIATED_UNIT_NAME[unit_name]
+
+            args["unit"] = Literal.string(unit_name)
         elif isinstance(unit, Week):
             unit.set("this", Literal.string(unit.this.name.upper()))
 
@@ -5570,6 +5585,10 @@ class Extract(Func):
     arg_types = {"this": True, "expression": True}
 
 
+class Exists(Func, SubqueryPredicate):
+    arg_types = {"this": True, "expression": False}
+
+
 class Timestamp(Func):
     arg_types = {"this": False, "zone": False, "with_tz": False}
 
@@ -5746,8 +5765,14 @@ class Greatest(Func):
     is_var_len_args = True
 
 
+# Trino's `ON OVERFLOW TRUNCATE [filler_string] {WITH | WITHOUT} COUNT`
+# https://trino.io/docs/current/functions/aggregate.html#listagg
+class OverflowTruncateBehavior(Expression):
+    arg_types = {"this": False, "with_count": True}
+
+
 class GroupConcat(AggFunc):
-    arg_types = {"this": True, "separator": False}
+    arg_types = {"this": True, "separator": False, "on_overflow": False}
 
 
 class Hex(Func):
@@ -5947,6 +5972,11 @@ class JSONBContains(Binary, Func):
     _sql_names = ["JSONB_CONTAINS"]
 
 
+class JSONBExists(Func):
+    arg_types = {"this": True, "path": True}
+    _sql_names = ["JSONB_EXISTS"]
+
+
 class JSONExtract(Binary, Func):
     arg_types = {
         "this": True,
@@ -6025,6 +6055,7 @@ class Levenshtein(Func):
         "ins_cost": False,
         "del_cost": False,
         "sub_cost": False,
+        "max_dist": False,
     }
 
 
@@ -6116,6 +6147,10 @@ class MD5Digest(Func):
     _sql_names = ["MD5_DIGEST"]
 
 
+class Median(AggFunc):
+    pass
+
+
 class Min(AggFunc):
     arg_types = {"this": True, "expressions": False}
     is_var_len_args = True
@@ -6198,11 +6233,11 @@ class Reduce(Func):
 class RegexpExtract(Func):
     arg_types = {
         "this": True,
-        "expression": True,
-        "position": False,
-        "occurrence": False,
-        "parameters": False,
-        "group": False,
+        "expression": True,  # The pattern
+        "position": False,  # Only start searching the string from this index
+        "occurrence": False,  # Skip the first `occurence-1` matches
+        "parameters": False,  # Flags, eg "i" for case-insensitive
+        "group": False,  # Which group to return
     }