gss-bi-udfs 0.1.0__py3-none-any.whl → 0.1.2__py3-none-any.whl

tests/test_utils.py

@@ -0,0 +1,119 @@
+import os
+import unittest
+from types import SimpleNamespace
+from unittest.mock import MagicMock, patch
+
+from pyspark.sql.types import (
+    BooleanType,
+    DateType,
+    DecimalType,
+    DoubleType,
+    FloatType,
+    IntegerType,
+    LongType,
+    StringType,
+    TimestampType,
+)
+
+from gss_bi_udfs import utils
+
+
+class TestUtils(unittest.TestCase):
+    def test_get_env_uses_default_and_env_var(self):
+        with patch.dict(os.environ, {}, clear=True):
+            self.assertEqual(utils.get_env(), "dev")
+            self.assertEqual(utils.get_env(default="qa"), "qa")
+
+        with patch.dict(os.environ, {"ENV": "prod"}, clear=True):
+            self.assertEqual(utils.get_env(), "prod")
+
+    def test_get_env_catalog(self):
+        with patch.dict(os.environ, {"ENV": "pro"}, clear=True):
+            self.assertEqual(utils.get_env_catalog("fi_comunes"), "fi_comunes")
+        with patch.dict(os.environ, {"ENV": "dev"}, clear=True):
+            self.assertEqual(utils.get_env_catalog("fi_comunes"), "fi_comunes_dev")
+
+    def test_get_env_table_path(self):
+        with patch("gss_bi_udfs.utils.get_env_catalog", return_value="cat_dev"):
+            self.assertEqual(
+                utils.get_env_table_path("cat", "silver.dim_cliente"),
+                "cat_dev.silver.dim_cliente",
+            )
+
+    @patch("gss_bi_udfs.utils.get_env_catalog", return_value="cat_dev")
+    def test_get_schema_root_location(self, _mock_catalog):
+        spark = MagicMock()
+        df = MagicMock()
+        df.filter.return_value = df
+        df.select.return_value = df
+        df.collect.return_value = [["s3://bucket/root"]]
+        spark.sql.return_value = df
+
+        out = utils.get_schema_root_location(spark, "cat", "silver")
+
+        self.assertEqual(out, "s3://bucket/root")
+        spark.sql.assert_called_once_with("DESCRIBE SCHEMA EXTENDED cat_dev.silver")
+
+    def test_get_table_info_validations(self):
+        spark = MagicMock()
+        with self.assertRaises(ValueError):
+            utils.get_table_info(spark, full_table_name="solo.dos")
+        with self.assertRaises(ValueError):
+            utils.get_table_info(spark)
+
+    @patch("gss_bi_udfs.utils.get_schema_root_location", return_value="s3://bucket/root/silver")
+    @patch("gss_bi_udfs.utils.get_env_catalog", return_value="cat_dev")
+    def test_get_table_info_when_table_does_not_exist(self, _mock_env_catalog, _mock_root):
+        spark = MagicMock()
+        spark.catalog.tableExists.return_value = False
+
+        info = utils.get_table_info(spark, full_table_name="cat.silver.dim_cliente")
+
+        self.assertEqual(info["catalog"], "cat_dev")
+        self.assertEqual(info["schema"], "silver")
+        self.assertEqual(info["table"], "dim_cliente")
+        self.assertEqual(info["full_table_name"], "cat_dev.silver.dim_cliente")
+        self.assertEqual(info["path"], "s3://bucket/root/silver/dim_cliente")
+        self.assertFalse(info["exists"])
+
+    @patch("gss_bi_udfs.utils.get_schema_root_location", return_value="s3://bucket/root/silver")
+    @patch("gss_bi_udfs.utils.get_env_catalog", return_value="cat_dev")
+    def test_get_table_info_when_table_exists(self, _mock_env_catalog, _mock_root):
+        spark = MagicMock()
+        spark.catalog.tableExists.return_value = True
+        desc_df = MagicMock()
+        desc_df.filter.return_value = desc_df
+        desc_df.collect.return_value = [
+            SimpleNamespace(col_name="Location", data_type="s3://bucket/real/location"),
+            SimpleNamespace(col_name="Provider", data_type="delta"),
+            SimpleNamespace(col_name="Type", data_type="MANAGED"),
+        ]
+        spark.sql.return_value = desc_df
+
+        info = utils.get_table_info(spark, full_table_name="cat.silver.dim_cliente")
+
+        self.assertTrue(info["exists"])
+        self.assertEqual(info["path"], "s3://bucket/real/location")
+        self.assertEqual(info["provider"], "delta")
+        self.assertEqual(info["table_type"], "MANAGED")
+
+    def test_get_default_value_by_type_returns_column(self):
+        dtypes = [
+            IntegerType(),
+            LongType(),
+            DecimalType(10, 2),
+            DoubleType(),
+            FloatType(),
+            DateType(),
+            TimestampType(),
+            BooleanType(),
+            StringType(),
+        ]
+        for dtype in dtypes:
+            with self.subTest(dtype=dtype):
+                out = utils.get_default_value_by_type(dtype)
+                self.assertEqual(out.__class__.__name__, "Column")
+
+
+if __name__ == "__main__":
+    unittest.main()

workspace/prueba.py

@@ -0,0 +1,10 @@
+from pyspark.sql import SparkSession
+
+spark = SparkSession.getActiveSession()
+if spark is None:
+    spark = SparkSession.builder.appName("MiApp").getOrCreate()
+
+df = spark.range(1000 * 1000)
+print(df.count())
+
+spark.stop()

workspace/prueba_calculadora.py

@@ -0,0 +1,20 @@
+class Calculadora:
+    """Clase calculadora con operaciones básicas"""
+    
+    def sumar(self, a, b):
+        """Suma dos números"""
+        return a + b
+    
+    def restar(self, a, b):
+        """Resta dos números"""
+        return a - b
+    
+    def multiplicar(self, a, b):
+        """Multiplica dos números"""
+        return a * b
+    
+    def dividir(self, a, b):
+        """Divide dos números"""
+        if b == 0:
+            raise ValueError("No se puede dividir entre cero")
+        return a / b

workspace/tests/test_prueba_calculadora.py

@@ -0,0 +1,14 @@
+import pytest
+import sys
+sys.path.append("/workspace")
+
+from .prueba_calculadora import sumar, restar
+
+def test_sumar():
+    assert sumar(2, 3) == 5
+
+def test_restar():
+    assert restar(5, 3) == 2
+
+def test_sumar_negativos():
+    assert sumar(-1, -2) == -3

gss_bi_udfs-0.1.0.dist-info/RECORD

@@ -1,4 +0,0 @@
-gss_bi_udfs-0.1.0.dist-info/METADATA,sha256=ChQSSxNCSDO4zK3mupeVxs9X0rXov2NWzib4n0HAoaM,339
-gss_bi_udfs-0.1.0.dist-info/WHEEL,sha256=qELbo2s1Yzl39ZmrAibXA2jjPLUYfnVhUNTlyF1rq0Y,92
-gss_bi_udfs-0.1.0.dist-info/top_level.txt,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-gss_bi_udfs-0.1.0.dist-info/RECORD,,

gss_bi_udfs-0.1.0.dist-info/top_level.txt

@@ -1 +0,0 @@
-