Summary
This chapter explains how to filter rows with missing data by using the SQL WHERE … IS NULL clause to check for NULL values in a column.

General domain of usage
Database querying

If a column doesn't contain any data, it's considered `NULL`. To determine if a value is `NULL`, you don’t just check for equality. Instead, you use the `WHERE` clause with `IS NULL` in a `SELECT` statement to find rows where the column’s value is `NULL`.

The syntax for it looks like this: 

```sql
SELECT columns
FROM table 
WHERE column_name IS NULL;
```

Let's illustrate this with an example:




SELECT name
FROM country
WHERE region IS NULL;

import unittest
import pandas as pd
import re
class TestTask(unittest.TestCase):
    @classmethod
    def setUpClass(cls):
        import psycopg2
        cls.connection = psycopg2.connect(
            host="localhost",
            user="agent",
            password="qwerty",
            dbname="testdb",
            sslmode="disable"
        )
        cls.cursor = cls.connection.cursor()
        # Create schema and populate data
        cls.cursor.execute("DROP TABLE IF EXISTS country;")
        cls.cursor.execute("""
            CREATE TABLE country (
                name VARCHAR(100),
                capital VARCHAR(100),
                population INTEGER
            );
        """)
        # Insert sample data: some with population, some with NULL
        countries = [
            ("France", "Paris", 67000000),
            ("Narnia", "Cair Paravel", None),
            ("Wakanda", "Birnin Zana", None),
            ("USA", "Washington", 331000000),
            ("Atlantis", "Poseidonis", None),
            ("Japan", "Tokyo", 125800000)
        ]
        cls.cursor.executemany("INSERT INTO country (name, capital, population) VALUES (%s, %s, %s);", countries)
        cls.connection.commit()

    def test_correct_columns_selected(self):
        """Check that only 'name' and 'capital' columns are selected."""
        with open("query.sql", "r") as f:
            user_query = f.read()
        self.cursor.execute(user_query)
        columns = [desc[0] for desc in self.cursor.description]
        expected_cols = ["name", "capital"]
        condition = columns == expected_cols
        _dynamic_test(
            self,
            condition,
            "Query selects only 'name' and 'capital' columns.",
            f"Query columns are {columns}, expected {expected_cols}."
        )

    def test_returns_only_null_population(self):
        """Check that only rows with population IS NULL are returned."""
        with open("query.sql", "r") as f:
            user_query = f.read()
        self.cursor.execute(user_query)
        rows = self.cursor.fetchall()
        condition = True
        for row in rows:
            name, capital = row
            self.cursor.execute(
                "SELECT population FROM country WHERE name = %s AND capital = %s",
                (name, capital)
            )
            pop = self.cursor.fetchone()
            if pop is None or pop[0] is not None:
                condition = False
                break
        _dynamic_test(
            self,
            condition,
            "All returned rows have population IS NULL.",
            "Some returned rows do not have population IS NULL."
        )

    def test_expected_rows_present(self):
        """Check that all expected countries with NULL population are present."""
        self.cursor.execute(
            "SELECT name, capital FROM country WHERE population IS NULL ORDER BY name, capital"
        )
        expected = self.cursor.fetchall()
        with open("query.sql", "r") as f:
            user_query = f.read()
        self.cursor.execute(user_query)
        user_rows = self.cursor.fetchall()
        expected_set = set(expected)
        user_set = set(user_rows)
        missing = expected_set - user_set
        extra = user_set - expected_set
        condition = (missing == set())
        fail_msg = ""
        if missing:
            fail_msg += f"Missing expected rows: {missing}. "
        if extra:
            fail_msg += f"Found unexpected rows: {extra}."
        _dynamic_test(
            self,
            condition,
            "All expected rows with NULL population are present.",
            fail_msg.strip()
        )

    def test_query_uses_is_null(self):
        """Check that the query uses 'IS NULL' in the WHERE clause for population."""
        with open("query.sql", "r") as f:
            user_query = f.read().lower().replace("\\n", " ")
        condition = ("where" in user_query and "population" in user_query and "is null" in user_query)
        _dynamic_test(
            self,
            condition,
            "Query uses IS NULL for population in WHERE clause.",
            "Query does not use 'IS NULL' for population in WHERE clause."
        )

def _dynamic_test(test_case, condition, success_message, failure_message):
    if condition:
        test_case._testMethodName = success_message
        test_case.assertTrue(True, success_message)
    else:
        test_case._testMethodName = failure_message
        test_case.fail(failure_message)

def normalize_text(text):
    text = text.lower()
    text = re.sub(r"\\s{2,}", " ", text)
    text = re.sub(r"\\s*([,:?])\\s*", r"\\1 ", text)
    return text.strip()

if __name__ == "__main__":
    unittest.main()


test_solution.py

Gain a solid understanding of SQL essentials tailored specifically for business intelligence. You will master querying, filtering, and aggregating business datasets through hands-on tasks designed to mirror real-world professional scenarios.

Checking for NO Value

Checking for NO Value

Ratkaisu