SICP-Notes-Lecture 21 SQL I

Lecture 21 SQL I

These are my notes for SICP(Structure and Interpretation of Computer Programs). Hope they’ll be of some help to you.

Declarative Programming

Programming paradigms

• Up till now, we’ve been focused (primarily) on imperative programming.

• Impreactive program contain explicit instructions to tell the computer how to accomplish something. The interpreter then executes those instructions.

• Now, we’ll learn about declarative programming, where we can just tell the computer what we want, instead of how we want it done. The interpreter then figures out how to accomplish that.

• Declarative programs are often specialized to perform a specific task, because they allow for repretitive computation to be abstracted away and for the interpreter to optimize its execution.

SQL

SQL & Database Languages

SQL is an example of a (declarative) language with interacts with a database management systems (DBMS) in order to make data processing easier and faster.

It collects records into tables, or a collection of rows with a value for each column.

• A row has a value for each column.
• A column has a name and a type.
• A table has columns and rows.

Tables in SQL (Sqlite, a small SQL database engine)

CREATE TABLE cities AS
	SELECT 38 AS latitude, 122 AS longtitude, "Berkeley" AS name UNION
	SELECT 42,			   71,                "Cambridge"		 UNION
	SELECT 45,			   93,   			  "Minneapolis";

SELECT "west coast" AS region, name FROM cities WHERE longtitude >= 115 UNION
SELECT "other", 			   name FROM cities WHERE longtitude < 115;

Cities:

SQL Basics

The SQL Language varies across implementations but we will look at some shared concepts.

• A SELECT statement creates a new table, either from scratch or by taking information from an existing table
• A CREATE TABLE statement gives a global name to a table.
• Lots of other statements exist: DELETE, INSERT, UPDATE etc…
• Most of the important action is in the SELECT statement.

Selecting value lierals

• A SELECT statement always includes a comma-separated list of column descriptions.

• A column description is an expression, optionally followed by AS and a column name.

  • SELECT [expression] AS [name], [expression] AS [name], ... ;

• SELECTing literals CREATEs a one-row table.

• The UNION of two SELECT statements is a table containing the rows of both of their results.

Naming Tables

SQL is often used as an interactive language.

• The result of a SELECT statement is displayed to the user, but not stored.
• A CREATE TABLE statement gives the result a name.
  • CREATE TABLE [name] AS [SELECT statements];

CREATE TABLE parents AS
	SELECT "delano" AS parent,  "herbert" AS child UNION
	SELECT "abraham",			"barack"		   UNION
	SELECT "abraham",			"clinton"		   UNION
	SELECT "fillmore",			"abraham"		   UNION
	SELECT "fillmore",			"delano"		   UNION
	SELECT "eisenhower",		"fillmore";

Parents:

Selecting From Tables

SELECT statements project existing tables

• A SELECT statement can specify an input table using a FROM clause.
• A subset of the rows of the input table can be selected using a WHERE clause.
• Can declare the order of the ramaining rows using an ORDER BY clause. Otherwise, no order.
• Column descriptions determine how each input row is projected to a result row:
  • SELECT [columns] FROM [table] WHERE [condition] ORDER BY [order] [ASC/DESC] LIMIT[number];

sqlite> SELECT * FROM parents ORDER BY parents DESC; --Only use ASC/DESC IF there's ORDER BY
sqlite> SELECT child FROM parents WHERE parent = "abraham";
sqlite> SELECT parent FROM parents WHERE parent > child;
--WHERE and ORDER BY are optional

Arithmetic in SELECT statements

• In a SELECCT expression, column names evaluate to row values.
• Arithmetic expressions can combine row values and constants.

CREATE TABLE restaurant AS
	SELECT 101 AS ttable, 2 AS single, 2 AS couple UNION
	SELECT 102		   , 0			, 3			  UNION
	SELECT 103		   , 3			, 1;

sqlite> SELECT ttable, single + 2 * couple AS total From restaurant;

Given a table ints that describes how to sum powers of 2 from various integers.

CREATE TABLE ints AS
	SELECT "zero" AS word, 0 AS one, 0 AS two, 0 AS four, 0 AS eight UNION
	SELECT "one"		 , 1	   , 0		 , 0		, 0			 UNION
	SELECT "two"		 , 0	   , 2		 , 0		, 0			 UNION
	SELECT "three"		 , 1	   , 2		 , 0		, 0			 UNION
	SELECT "four"		 , 0	   , 0		 , 4		, 0			 UNION
	SELECT "five"		 , 1	   , 0		 , 4		, 0			 UNION
	SELECT "six"		 , 0	   , 2		 , 4		, 0			 UNION
	SELECT "seven"		 , 1	   , 2		 , 4		, 0			 UNION
	SELECT "eight"		 , 0	   , 0		 , 0		, 8			 UNION
	SELECT "nine"		 , 1	   , 0		 , 0		, 8;

1. Write a SELECT statement for a two-column table of the word and the value for each integer.

sqlite> SELECT word, one + two + four + eight AS value FROM ints; 

2. Write a SELECT statement for the word names of the powers of two.

sqlite> SELECT word  FROM ints WHERE one + two + four + eight = 1 OR one + two + four + eight = 2 OR one + two + four + eight = 4 OR one + two + four + eight = 8;
--The code above has syntax error but can convey the logic.

Joinning Tables

An example:

CREATE TABLE dogs AS
	SELECT "abraham" AS name, "long" AS fur UNION
	SELECT "barack"			, "short" 		UNION
	SELECT "clinton"		, "long"		UNION
	SELECT "delano"			, "long"		UNION
	SELECT "esisenhower"	, "short"		UNION
	SELECT "fillmore"		, "curly"		UNION
	SELECT "grover"			, "short"		UNION
	SELECT "herbert"		, "curly";

How can we select the parents of curly furred dogs?

Joining tables

• Two tables A & B are joined by a comma to yield all combinations of a row from A & a row from B.

  SELECT * FROM parents, dogs;
  

• Selects all combinations of rows from both tables. We only want the rows for curly haired dogs.

  SELECT * FROM parents, dogs WHERE fur = "curly";
  

• This filters the 56 rows to now only have rows where the fur is curly. But this has rows that have nothing to do with each other. We only care about rows where the two dogs match.

  SELECT * FROM parents, dogs WHERE child = name AND fur = "curly";
  

• The condition on which the tables are joined on is called the join condition.

  SELECT parent FROM parents, dogs WHERE child = name AND fur = "curly";
  

Joining a table with itself

• Two tables may share a column name; dot expressions and aliases disambiguate column values.

SELECT [columns] FROM [table] WHERE [condition] ORDER BY [order];

[table] is a comma-separated list of table names with optional aliases.

Select all pairs of siblings. No duplicates.

Aliasing

SELECT * FROM parents, parents;
--This doesn't work because the tables share a column. Lets fix that!

SELECT * FROM parents AS a, parents AS b;
--This works because SQL can tell the columns in the two tables apart.
--Let's now only keep rows where the children share a parent.

SELECT * FROM parents AS a, parents AS b WHERE a.parent = b.parent;
--We need to get rid of duplicates because pairs of siblings appear twice.

--We can do this by enforcing an arbitrary ordering, a.child < b.child alphabetically.
--Then we get the two columns we want.
SELECT a.child AS first, b.child AS second FROM parents AS a, parents AS b WHERE a.parent = b.parent AND a.child < b.child;

String Expressions

String values can be combined to form longer strings.

sqlite> SELECT "hello," || " world";
hello, world
sqlite> SELECT name || " dog" FROM dogs;
abraham dog
barack dog
clinton dog
delano dog
eisenhower dog
fillmore dog
grover dog
herbert dog