fix(teradata): transpile SAMPLE row counts and fractions [CLAUDE]

sqlglot/generator.py

@@ -2566,6 +2566,10 @@ def tablesample_sql(
         expression: exp.TableSample,
         tablesample_keyword: str | None = None,
     ) -> str:
+        if expression.expressions:
+            self.unsupported("Multiple sample sets are not supported")
+            return ""
+
         method = self.sql(expression, "method")
         method = f"{method} " if method and self.TABLESAMPLE_WITH_METHOD else ""
         numerator = self.sql(expression, "bucket_numerator")

sqlglot/generators/teradata.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import typing as t
+from decimal import Decimal
 
 from sqlglot import exp, generator, transforms
 from sqlglot.dialects.dialect import (
@@ -116,7 +117,16 @@ def tablesample_sql(
         expression: exp.TableSample,
         tablesample_keyword: str | None = None,
     ) -> str:
-        return f"{self.sql(expression, 'this')} SAMPLE {self.expressions(expression)}"
+        sample = expression.args.get("size")
+        percent = expression.args.get("percent")
+        if percent and percent.is_number:
+            fraction = Decimal(percent.to_py()).scaleb(-2)
+            # 100 percent keeps its decimal point, otherwise it'd be read back as a row count
+            sample = exp.Literal.number(
+                format(fraction.normalize() if fraction < 1 else fraction, "f")
+            )
+
+        return f" SAMPLE {self.sql(sample) or self.expressions(expression)}"
 
     def window_sql(self, expression: exp.Window) -> str:
         spec = expression.args.get("spec")

sqlglot/parsers/teradata.py

@@ -146,6 +146,21 @@ def _parse_update(self) -> exp.Update:
             )
         )
 
+    def _parse_table_sample(self, as_modifier: bool = False) -> exp.TableSample | None:
+        sample = super()._parse_table_sample(as_modifier=as_modifier)
+
+        # A single integer sample is a row count, while a fraction is a proportion of rows
+        # https://docs.teradata.com/r/Enterprise_IntelliFlex_VMware/SQL-Data-Manipulation-Language/SELECT-Statements/SAMPLE-Clause/SAMPLE-Clause-Syntax
+        if sample and len(sample.expressions) == 1 and sample.expressions[0].is_number:
+            num = sample.expressions[0].pop()
+            if num.is_int:
+                sample.set("size", num)
+            else:
+                percent = format(num.to_py().scaleb(2).normalize(), "f")
+                sample.set("percent", exp.Literal.number(percent))
+
+        return sample
+
     def _parse_rangen(self):
         this = self._parse_id_var()
         self._match(TokenType.BETWEEN)

tests/dialects/test_teradata.py

@@ -1,4 +1,4 @@
-from sqlglot import exp
+from sqlglot import ErrorLevel, UnsupportedError, exp
 from tests.dialects.test_dialect import Validator
 
 
@@ -20,6 +20,34 @@ def test_teradata(self):
             "SELECT * FROM tbl SAMPLE 0.33, .25, .1",
             "SELECT * FROM tbl SAMPLE 0.33, 0.25, 0.1",
         )
+        self.validate_all(
+            "SELECT * FROM t SAMPLE 100",
+            write={
+                "teradata": "SELECT * FROM t SAMPLE 100",
+                "snowflake": "SELECT * FROM t TABLESAMPLE (100 ROWS)",
+            },
+        )
+        self.validate_all(
+            "SELECT * FROM t SAMPLE 0.25",
+            read={
+                "snowflake": "SELECT * FROM t TABLESAMPLE (25)",
+            },
+            write={
+                "teradata": "SELECT * FROM t SAMPLE 0.25",
+                "snowflake": "SELECT * FROM t TABLESAMPLE (25)",
+            },
+        )
+        self.validate_all(
+            "SELECT * FROM t SAMPLE 0.6, 0.25",
+            write={
+                "teradata": "SELECT * FROM t SAMPLE 0.6, 0.25",
+                "snowflake": "SELECT * FROM t",
+            },
+        )
+        with self.assertRaises(UnsupportedError):
+            self.parse_one("SELECT * FROM t SAMPLE 10, 20").sql(
+                "snowflake", unsupported_level=ErrorLevel.RAISE
+            )
 
         self.validate_all(
             "DATABASE tduser",