feat: add bigframes.bigquery.to_json_string

Author: chelsea-lin

bigframes/bigquery/__init__.py

@@ -50,6 +50,7 @@
     json_value,
     json_value_array,
     parse_json,
+    to_json_string,
 )
 from bigframes.bigquery._operations.search import create_vector_index, vector_search
 from bigframes.bigquery._operations.sql import sql_scalar
@@ -87,6 +88,7 @@
     json_value,
     json_value_array,
     parse_json,
+    to_json_string,
     # search ops
     create_vector_index,
     vector_search,

bigframes/bigquery/_operations/json.py

@@ -430,6 +430,40 @@ def json_value_array(
     return input._apply_unary_op(ops.JSONValueArray(json_path=json_path))
 
 
+def to_json_string(
+    input: series.Series,
+) -> series.Series:
+    """Converts a series to a JSON-formatted STRING value.
+
+    **Examples:**
+
+        >>> import bigframes.pandas as bpd
+        >>> import bigframes.bigquery as bbq
+        >>> bpd.options.display.progress_bar = None
+
+        >>> s = bpd.Series([1, 2, 3])
+        >>> bbq.to_json_string(s)
+        0    1
+        1    2
+        2    3
+        dtype: string
+
+        >>> s = bpd.Series([{"int": 1, "str": "pandas"}, {"int": 2, "str": "numpy"}])
+        >>> bbq.to_json_string(s)
+        0    {"int":1,"str":"pandas"}
+        1     {"int":2,"str":"numpy"}
+        dtype: string
+
+    Args:
+        input (bigframes.series.Series):
+            The Series to be converted.
+
+    Returns:
+        bigframes.series.Series: A new Series with the JSON-formatted STRING value.
+    """
+    return input._apply_unary_op(ops.ToJSONString())
+
+
 @utils.preview(name="The JSON-related API `parse_json`")
 def parse_json(
     input: series.Series,

bigframes/core/compile/ibis_compiler/scalar_op_registry.py

@@ -2068,9 +2068,7 @@ def json_extract_string_array(  # type: ignore[empty-body]
 
 
 @ibis_udf.scalar.builtin(name="to_json_string")
-def to_json_string(  # type: ignore[empty-body]
-    json_obj: ibis_dtypes.JSON,
-) -> ibis_dtypes.String:
+def to_json_string(json_obj) -> ibis_dtypes.String:  # type: ignore[empty-body]
     """Convert JSON to STRING."""
 
 

bigframes/operations/json_ops.py

@@ -108,10 +108,10 @@ class ToJSONString(base_ops.UnaryOp):
 
     def output_type(self, *input_types):
         input_type = input_types[0]
-        if not dtypes.is_json_like(input_type):
+        if not dtypes.is_json_encoding_type(input_type):
             raise TypeError(
-                "Input type must be a valid JSON object or JSON-formatted string type."
-                + f" Received type: {input_type}"
+                "The value to be assigned must be a type that can be encoded as JSON."
+                + f"Received type: {input_type}"
             )
         return dtypes.STRING_DTYPE
 

tests/system/small/bigquery/test_json.py

@@ -384,3 +384,28 @@ def test_parse_json_w_invalid_series_type():
     s = bpd.Series([1, 2])
     with pytest.raises(TypeError):
         bbq.parse_json(s)
+
+
+def test_to_json_string_from_int():
+    s = bpd.Series([1, 2, None, 3])
+    actual = bbq.to_json_string(s)
+    expected = bpd.Series(["1", "2", "null", "3"], dtype=dtypes.STRING_DTYPE)
+    pd.testing.assert_series_equal(actual.to_pandas(), expected.to_pandas())
+
+
+def test_to_json_string_from_struct():
+    s = bpd.Series(
+        [
+            {"version": 1, "project": "pandas"},
+            {"version": 2, "project": "numpy"},
+        ]
+    )
+    assert dtypes.is_struct_like(s.dtype)
+
+    actual = bbq.to_json_string(s)
+    expected = bpd.Series(
+        ['{"project":"pandas","version":1}', '{"project":"numpy","version":2}'],
+        dtype=dtypes.STRING_DTYPE,
+    )
+
+    pd.testing.assert_series_equal(actual.to_pandas(), expected.to_pandas())