FHWN DB Test,DQL,Views

Author: Mate Hegyhati

FHWN/DB/2025_10_03_Test_DQL_Views/DQL_DDL.md

@@ -0,0 +1,269 @@
+# Play along
+
+For testing online, visit [this site](https://datawithdev.com/sql-playground/) and import the data from [the sqlite DB](/athlete_data_after2000.db).
+
+Or use [DB Browsaer for SQLite](https://sqlitebrowser.org/) locally.
+
+Data source: [Kaggle](https://www.kaggle.com/datasets/krishd123/olympics-legacy-1896-2020)
+
+# DQL
+
+## Basic commands
+
+### Get all the data
+```sql
+SELECT * FROM athlete_results;
+```
+
+### Get only specific columnns
+Name and team for all of the results:
+```sql
+SELECT Name,Team FROM athlete_results;
+```
+
+> [!NOTE] 
+> Keywords can be in any case, `SELECT`, `select` and `sEleCt` does the same.
+> But stick to one (of the first 2).
+
+### Filter by condition
+Get the name and sport for each Austrian athlete:
+```sql
+SELECT Name,Sport FROM athlete_results WHERE NOC = 'AUT';
+```
+> [!IMPORTANT]
+> Double qutes are **NOT** used for string literals, they have to be used for table/column names if they contain spaces.
+
+> [!NOTE]
+> `=` is the "is equal" operator, not `==`. 
+> But it is also accepted by a lot of DBMSs.
+
+
+### Filter by multiple conditions
+Get the name, exact event and medal of Austrian athletes who ended up on the podium:
+
+```sql
+SELECT Name, Event, Medal FROM athlete_results WHERE NOC = 'AUT' AND Medal != NULL;
+```
+
+> [!Caution]
+> 0 results, because `NULL` is special. 
+> 
+> `anything = NULL` -> `false`
+>
+> `anything != NULL` -> `false`
+> 
+> Even:
+> 
+> `NULL = NULL` -> `false`
+> 
+> `NULL != NULL` -> `false`
+
+Always use `IS NULL` or `IS NOT NULL`:
+
+```sql
+SELECT Name, Event, Medal FROM athlete_results WHERE NOC = 'AUT' AND Medal IS NOT NULL;
+```
+
+> [!NOTE]
+> SQLite is lenient in many things, `IS` can be left out for example, but not proper SQL.
+
+
+
+### Filter comparing columns
+List where Country and Team differ:
+```sql
+SELECT Country,Team FROM athlete_results WHERE Country != Team;
+```
+
+### Remove duplicates
+Same, but list each only once:
+```sql
+SELECT DISTINCT Country,Team FROM athlete_results WHERE Country != Team;
+```
+
+> [!Important]
+> `DISTINCT` applies to the whole `Country,Team` tuple, so the same `Country` could appear twice if it had different `Team`s.
+
+### Get the number of results
+How many Gold medals did Austrian athletes won?
+```sql
+SELECT COUNT(*) FROM athlete_results where NOC = 'AUT' AND Medal = 'Gold';
+```
+> [!Note]
+> `COUNT()` is also often accepted, again, not proper.
+
+
+How many Austrian athletes won Gold medals?
+```sql
+SELECT COUNT(DISTINCT Name) FROM athlete_results where NOC = 'AUT' AND Medal = 'Gold';
+```
+
+### Order results
+Order the countries by name:
+```sql
+SELECT DISTINCT Country FROM athlete_results ORDER BY Country;
+```
+
+## Grouping and nesting
+
+
+
+### Group by attribute
+
+Group Austrian athletes by Sport:
+```sql
+SELECT Name, Sport, Year FROM athlete_results WHERE NOC = 'AUT' GROUP BY Sport;
+```
+Which values are selected for `Name` and `Year` for each `Sport` group?!
+
+> [!TIP]
+> When using `GROUP BY`, it is a code smell if any non-groupping column remains as is without any aggregate function.
+
+Count how many Gold medals each Country had:
+```sql
+SELECT Country, COUNT(Medal) FROM athlete_results WHERE Medal = 'Gold' GROUP BY Country;
+```
+
+Little bit nicer:
+```sql
+SELECT Country, COUNT(*) AS "Gold Medals" FROM athlete_results WHERE Medal = 'Gold' GROUP BY Country;
+```
+
+And sort it then:
+```SQL
+SELECT Country, COUNT(*) AS "Gold Medals" FROM athlete_results WHERE Medal = 'Gold' GROUP BY Country ORDER BY "Gold Medals";
+```
+
+Descending...
+```SQL
+SELECT Country, COUNT(*) AS "Gold Medals" FROM athlete_results WHERE Medal = 'Gold' GROUP BY Country ORDER BY "Gold Medals" DESC;
+```
+
+Only top 20:
+```sql
+SELECT Country, COUNT(*) AS "Gold Medals" FROM athlete_results WHERE Medal = 'Gold' GROUP BY Country ORDER BY "Gold Medals" DESC LIMIT 20;
+```
+
+Probably it is time to break it up to multiple lines for readibility:
+```sql
+SELECT 
+    Country, 
+    COUNT(*) AS "Gold Medals" 
+FROM athlete_results 
+WHERE Medal = 'Gold' 
+GROUP BY Country 
+ORDER BY "Gold Medals" DESC 
+LIMIT 20;
+```
+
+### Nesting
+Sometimes question cannot (easily) be answered in one go, an intermediate table is needed/useful.
+
+What is the average number of Gold medals won by a Country in each event it participates?
+
+Distinct participation of countries easy:
+```sql
+SELECT DISTINCT Country, Year, FROM athlete_results;
+```
+
+Count the medals:
+```sql
+SELECT 
+    Country, 
+    Year, 
+    COUNT(Medal) AS Gold
+FROM athlete_results
+GROUP BY Country, Year;
+```
+
+This result as an "intermediate table" can be used for a subsequent query:
+```sql
+SELECT Country, AVG(Gold) AS "Average gold"
+FROM (
+    SELECT 
+        Country, 
+        Year, 
+        COUNT(Medal) AS Gold
+    FROM athlete_results
+    GROUP BY Country, Year;
+) t
+GROUP BY Country;
+ORDER BY "Average gold"
+```
+
+
+## Practice questions:
+ - How many countries won a gold medal in the 2020 Olympics? 
+ - Who participated at the most olympic games?
+ - Who won the most Gold medals in a single event?
+ - Was there an athlete who changed citizenship? 
+ - Who are the two athletes that shared the podium the most times? 
+
+# DDL
+
+## Views
+
+### Create (virtual) view
+```sql
+CREATE VIEW most_participations AS
+SELECT Name, COUNT(*) AS Participation FROM (
+    SELECT DISTINCT Name, Year, Season 
+    FROM athlete_results 
+) t
+GROUP BY Name
+ORDER BY Participation DESC
+```
+
+> [If it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck.](https://en.wikipedia.org/wiki/Duck_test)
+
+Views...
+ - ✅ look like tables
+ - ✅ can be queried like a table
+ - ❌ cannot be altered (DML) as tables (usually).
+ - ❌ are not stored like tables
+
+Views are not tables.
+The results of the "construction query" is not stored, only the "recipe query". 
+When querying a view, the recipe query is executed first to get the view.
+
+### Querying View
+Just like a table:
+```sql
+SELECT * from most_participations LIMIT 10;
+```
+
+### Create materialized View
+Kinda like tables, data is cached also, but needs to be updated (mostly manually).
+
+Not standard!
+
+```sql
+CREATE MATERIALIZED VIEW most_participations_materialized AS
+SELECT Name, COUNT(*) AS Participation FROM (
+    SELECT DISTINCT Name, Year, Season 
+    FROM athlete_results 
+) t
+GROUP BY Name
+ORDER BY Participation DESC
+```
+
+### Virtual vs. Material views
+
+| | Virtual | Material|
+| --- | --- | --- |
+| Pro | small memory footrpint, always-up-to-date | faster to query, saves CPU | 
+| Con | extra computation for every query | not up-to-date necessarily, additional memory need |
+
+## Dropping
+
+### Drop entire table or view
+```sql
+DROP VIEW most_participations;
+```
+![SQL Injection](Injection.PNG)
+
+
+### Delete column/attribute
+```sql
+ALTER TABLE athlete_results DROP COLUMN A;
+```

FHWN/DB/2025_10_03_Test_DQL_Views/Injection.PNG

@@ -0,0 +1,42 @@
+# Test
+
+## Description
+
+Draw the ER diagram for the following use case and write the SQL statements to create the corresponding schema.
+
+> There are many airports around the world and each has its 3-letter identifier. For example VIE for Vienna - Schwechat International Airport. About each airport we know the number of both runways and gates, moreover whether it is international or not. Each airport is located in a country, that has a unique 2 digit (e.g., AT) and 3 digit (AUT) code. We also want to store the population of each country, and whether they are in Schengen or not. Finally, we would like to store, which airports are connected by at least one carrier.
+
+`NOT NULL` and `ON DELETE` statements can be omitted.
+
+## Solution
+
+**One** good solution, there are more.
+
+### ERD
+
+![ERD of test](./test_erd.svg)
+
+### SQL DDL
+
+```sql
+CREATE TABLE Airports (
+    iata VARCHAR(3) PRIMARY KEY,
+    runway_count INT,
+    gate_count INT,
+    is_international BOOLEAN,
+    country VARCHAR(3) REFERENCES Countries(alpha3)
+);
+
+CREATE TABLE Countries (
+    alpha2 VARCHAR(2) UNIQUE,
+    alpha3 VARCHAR(3) PRIMARY KEY,
+    population INT,
+    in_schengen BOOLEAN
+);
+
+CREATE TABLE Connections (
+    iata1 VARCHAR(3) REFERENCES Airports(iata),
+    iata2 VARCHAR(3) REFERENCES Airports(iata),
+    PRIMARY KEY(iata1,iata2)
+);
+```