devkuma – SQLite

SQLite | Installing SQLite

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

SQLite manages each database as a single file without using a database server. Because it does not use a database server, there is no separate application that must run continuously as a server. To use SQLite from the command line, you use a separate command-line interface program. To use it from PHP, Ruby on Rails, and similar environments, you need a DLL. This section explains how to download and install the SQLite command-line program and DLL.

SQLite | Installing SQLite | Downloading and Installing the SQLite Command-Line Tool

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains how to download and install the SQLite command-line program so that you can use SQLite from the command line. As of October 2019, the latest version was SQLite 3.30.1 (2019-10-11).

Download SQLite

Open the official SQLite website in a browser.

https://www.sqlite.org/

Click the “Download” menu at the top of the page.

Click the tool binary for your operating system, such as a file named in the sqlite-tools-xxx format, download it, and save it wherever you prefer.

Install the SQLite Command-Line Program

The downloaded file is a compressed archive. Installation is complete once you extract it to any directory.

$ ls -al
total 5016
drwxr-xr-x@  5 kimkc  staff      170 10 11 18:32 .
drwxr-xr-x  41 kimkc  staff     1394 10 14 23:41 ..
-rwxr-xr-x@  1 kimkc  staff   691768 10 11 18:31 sqldiff
-rwxr-xr-x@  1 kimkc  staff  1152260 10 11 18:32 sqlite3
-rwxr-xr-x@  1 kimkc  staff   719796 10 11 18:31 sqlite3_analyzer

The archive contains three files: sqlite3, the command-line tool; sqldiff, which reports differences between two databases; and sqlite3_analyzer, which displays database analysis reports.

SQLite | Installing SQLite | Testing the SQLite Command-Line Tool

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page shows how to try the downloaded SQLite command-line tool and confirm that it works.

Create a Database and Table with the Command-Line Tool

First, start a command prompt. Because PATH has not been configured separately, move to the directory where the sqlite3 executable is located.

Create a database. To create a database with the command-line tool, use the following format.

sqlite3 database-name

When you specify a database name and run the sqlite3 program, SQLite connects to the database with that name if it already exists. If the database with that name does not exist, SQLite creates a new database and then connects to it.

The database name can be anything, but a file is created with the specified database name. For example, you can use sampledb.sqlite3 or sampledb.db. A name without an extension, such as sampledb, is also fine. Here, use sample.sqlite3.

sqlite3 sample.sqlite3

$ sqlite3 sample.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>

A new database named sample.sqlite3 has been created and connected. However, the actual file is created only after you create an object such as a table in the database. While connected to SQLite, the sqlite> prompt is displayed.

Next, create one table in the database. Run the following statement.

create table user(id, name);

$ sqlite3 sample.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> create table user(id, name);
sqlite>

The table has been created. This example created a database and then created a table in that database.

To close the connection to the database, enter .exit.

.exit

$ sqlite3 sample.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> create table user(id, name);
sqlite> .exit

The connection to the database has been closed.

The File Where the Database Is Stored

After you create a database, data is stored in the database file when you create an object such as a table. Because no location was specified separately, the database file is created in the directory where the sqlite3 executable is located.

$ ls
sample.sqlite3   sqldiff          sqlite3          sqlite3_analyzer

You can see that a new file named sample.sqlite3 has been created. This file stores the data for the sample.sqlite3 database you just created. In this way, SQLite creates and manages one file for each database, although in some cases it may use multiple files.

SQLite | Installing SQLite | Downloading sqlite3.dll

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

SQLite can be used from the command line, but it is more often used from other applications such as PHP and Rails. To use SQLite from an application, you need sqlite3.dll. This page explains how to download the sqlite3.dll file.

Download sqlite3.dll

To download sqlite3.dll, open the following URL.

https://www.sqlite.org/download.html

Open the official SQLite page in a browser.

In a 64-bit Windows environment, click the link named sqlite-dll-win64-x64-3300100.zip in the “Precompiled Binaries for Windows” block, download it, and save it wherever you prefer.

Copy sqlite3.dll to a Folder Configured in PATH

When you extract the downloaded sqlite-dll-win64-x64-3300100.zip archive, it contains two files: sqlite3.dll and sqlite3.def. Save the sqlite3.dll file to a folder configured in PATH. For example, you can place it in C:\Windows\System32.

sqlite3.def sqlite3.dll

In fact, when you install PHP, the DLL required to use SQLite is also installed with it. Therefore, downloading it separately is not always required.

SQLite | Installing SQLite | Referencing the SQLite Documentation

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains how to reference the SQLite documentation. You can read it online or download it for local use.

Referencing the Online Documentation

To reference the online documentation, click “Documentation” in the menu at the top of the official SQLite page, or open the following link.

https://www.sqlite.org/docs.html

For example, expand “Programming Interfaces” and click the “SQL Syntax” menu.

You can view explanations of the SQL features available in SQLite.

You can also reference many other documents about SQLite.

Downloading the Documentation

If you want to download the documentation locally, click the “Download” menu at the top of the page, or open the following link.

https://www.sqlite.org/download.html

Click the “sqlite-doc-3300100.zip” link in the “Documentation” block to download the documentation.

After the download is complete, extract the compressed file. Many HTML files will appear, as shown below.

Find and open the index.html file.

The same screen as the official website page is displayed. Click the “Documentation” menu at the top of the screen, and you can reference the documentation offline as well.

SQLite

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This section explains how to use the SQLite database. SQLite usually runs inside a single application rather than as a database server. It is easy to install and very compact, so it is often distributed together with applications. Here, each topic walks through how to use SQLite to create databases and tables, add data, and query data.

SQLite | SQLite Basic Knowledge

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This section explains the basic knowledge you should understand when using SQLite.

SQLite | SQLite Basic Knowledge | How to Enter SQL Statements and Commands in the Command-Line Tool

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

In SQLite, you can use the command-line tool to create databases, search data, and perform other operations. This page explains how to enter commands when using the SQLite command-line tool.

Running Commands

To run the SQLite command-line tool, start a command prompt, move to the directory where sqlite3 is installed, and run the following command. The tool starts and connects to the database specified as the argument.

sqlite3 database-name

This time, run the following command to connect to the sample.sqlite3 database created earlier. Even if the previously created file was deleted or does not exist, the connection command itself still works.

$ sqlite3 sample.sqlite3 
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>

In the command-line tool, you can run SQLite commands that configure the command-line tool, as well as arbitrary SQL statements.

Try running the SQLite command .show. Enter the command, add any required arguments if needed, and press the Enter key to execute it.

$ sqlite3 sample.sqlite3 
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .show
        echo: off
         eqp: off
     explain: auto
     headers: off
        mode: list
   nullvalue: ""
      output: stdout
colseparator: "|"
rowseparator: "\n"
       stats: off
       width: 
    filename: sample.sqlite3
sqlite>

Running SQL Statements

Next, run a simple SQL statement. For SQL statements, everything up to the final semicolon (;) is treated as one SQL statement. After entering the semicolon, press Enter to execute it.

sqlite> create table user(id, name);
sqlite> insert into user(id, name) values (1, 'devkuma');
sqlite> insert into user(id, name) values (2, 'araikuma');
sqlite> select * from user;
1|devkuma
2|araikuma
sqlite>

If the user table created earlier already exists, you do not need to run the first line.

Splitting a Long SQL Statement Across Multiple Lines

A single SQL statement can become long. If you try to enter the entire statement at once, it can be difficult to type or hard to understand how far you have entered.

sqlite> select id as user_id, name as user_name from user where id = 1 and name = 'devkuma' order by name desc;
1|devkuma
sqlite>

When a statement is long, the command-line tool also allows you to split it across multiple lines. Press Enter while entering the statement.

sqlite> select id as user_id, name as user_name 
   ...>

SQLite treats everything up to the final semicolon (;) as one statement. If you press Enter before entering the semicolon, SQLite determines that the statement is still being entered and lets you continue.

Now enter the rest of the statement. Press Enter at appropriate points as shown below.

sqlite> select id as user_id, name as user_name 
   ...> from user
   ...> where id = 1 and name = 'devkuma'
   ...> order by name desc
   ...>

You do not need to add a space before pressing Enter. When you press Enter in the middle of a command, the text before and after it is automatically treated as separate words.

For SQL statements, when you enter a line that contains ; and press Enter, SQLite determines that the statement input is complete and executes the SQL statement.

sqlite> select id as user_id, name as user_name 
   ...> from user
   ...> where id = 1 and name = 'devkuma'
   ...> order by name desc
   ...> ;
1|devkuma
sqlite>

Entering the statement on one line and splitting it across multiple lines both produce the same result, so use whichever style fits the situation.

–

This page explained how to enter commands and SQL statements in the SQLite command-line tool.

SQLite | SQLite Basic Knowledge | Notes on Keywords Defined in SQLite

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

SQLite defines many keywords. These keywords cannot normally be used as table names, column names, and similar identifiers, but you can use them if you write them in an appropriate form. This page explains points to note about keywords defined in SQLite and lists those keywords.

How to Use Keywords as Identifiers

SQLite defines many keywords. The list is shown later, but words such as TABLE and SELECT are keywords. Keywords are also called reserved words.

Table names and database names are called identifiers. Identifiers can be specified by combining letters, numbers, and similar characters. For example, names such as booktable and name are valid. However, SQLite keywords cannot be used directly as identifiers. For example, an error occurs if you try to use select as a table name.

create table select(id, name);

sqlite> create table select(id, name);
Error: near "select": syntax error
sqlite>

It is better to avoid using keywords as table names, database names, and similar identifiers whenever possible. However, if you need to use a keyword as an identifier for some reason, use one of the following four forms.

'keyword'
"keyword"
[keyword]
`keyword`

When a keyword is enclosed in single quotes ('), it is treated as a string value. If a single-quoted string value is written where an identifier is expected, SQLite treats it as an identifier, so you can specify it as follows.

create table 'select'(id, name);

sqlite> create table 'select'(id, name);
sqlite>

When a keyword is enclosed in double quotes ("), square brackets ([]), or backticks (`), it is treated as an identifier. Also, if you write an identifier enclosed in double quotes where a string should be written, it is treated as a string.

create table "select"(id, name);

sqlite> create table "select"(id, name);
Error: table "select" already exists
sqlite> drop table 'select';
sqlite> create table "select"(id, name);
sqlite>

Square brackets are the style used by Access and SQL Server, while backticks are the style used by MySQL. These two styles are provided for compatibility with those databases, so in general, use single quotes or double quotes.

Keyword List

The following are the keywords defined by SQLite. This list is published on the official site below.

SQLite Keywords

ABORT	ACTION	ADD	AFTER	ALL
ALTER	ANALYZE	AND	AS	ASC
ATTACH	AUTOINCREMENT	BEFORE	BEGIN	BETWEEN
BY	CASCADE	CASE	CAST	CHECK
COLLATE	COLUMN	COMMIT	CONFLICT	CONSTRAINT
CREATE	CROSS	CURRENT	CURRENT_DATE	CURRENT_TIME
CURRENT_TIMESTAMP	DATABASE	DEFAULT	DEFERRABLE	DEFERRED
DELETE	DESC	DETACH	DISTINCT	DO
DROP	EACH	ELSE	END	ESCAPE
EXCEPT	EXCLUSIVE	EXISTS	EXPLAIN	FAIL
FILTER	FOLLOWING	FOR	FOREIGN	FROM
FULL	GLOB	GROUP	HAVING	IF
IGNORE	IMMEDIATE	IN	INDEX	INDEXED
INITIALLY	INNER	INSERT	INSTEAD	INTERSECT
INTO	IS	ISNULL	JOIN	KEY
LEFT	LIKE	LIMIT	MATCH	NATURAL
NO	NOT	NOTHING	NOTNULL	NULL
OF	OFFSET	ON	OR	ORDER
OUTER	OVER	PARTITION	PLAN	PRAGMA
PRECEDING	PRIMARY	QUERY	RAISE	RANGE RECURSIVE
REFERENCES	REGEXP	REINDEX	RELEASE	RENAME
REPLACE	RESTRICT	RIGHT	ROLLBACK	ROW
ROWS	SAVEPOINT	SELECT	SET	TABLE
TEMP	TEMPORARY	THEN	TO	TRANSACTION
TRIGGER	UNBOUNDED	UNION	UNIQUE	UPDATE
USING	VACUUM	VALUES	VIEW	VIRTUAL
WHEN	WHERE	WINDOW	WITH	WITHOUT

Some of these keywords can be used as identifiers even if they are not enclosed in single quotes, double quotes, or similar delimiters. However, using keywords directly as identifiers makes them harder to understand, so it is not recommended.

If an unknown error occurs when creating a table or similar object, check whether the table name or column name is a keyword.

SQLite | SQLite Basic Knowledge | Writing Comments in SQL Statements

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains the formats used to write comments in SQL statements.

Writing Comments

When you want to leave notes or explanations in SQL statements, SQLite provides the following two ways to add comments to SQL statements.

-- comment

/* comment */

Comments are ignored when SQL statements are executed, so they do not affect the execution result.

When you write a comment in the -- comment format, the text from -- to the end of the line becomes a comment. For example, write it as follows.

select * from user; -- list

sqlite> select * from user; -- list
1|devkuma
2|araikuma
sqlite>

When you write a comment in the /* comment */ format, the text from /* to */ becomes a comment.

select * from /* comment */ user;

sqlite> select * from /* comment */ user;
1|devkuma
2|araikuma
sqlite>

You may not write comments very often in the command-line tool, but comments are useful when you write multiple SQL statements in an external file and then import and execute them in SQLite.

SQLite | Database

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This section explains how to create a database with SQLite3 and how to delete a database that you have created.

SQLite | Database | Creating and Connecting to a Database

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains how to create a new SQLite database and how to connect to a database that has already been created.

Creating a Database

To create a new database, enter the following command at the command prompt.

sqlite3 database-name

For example, suppose you want to create a database named myfriend.sqlite3. Start a command prompt, move to the directory that contains the sqlite3 file, and run the following command.

$ sqlite3 myfriend.sqlite3

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>

If the myfriend.sqlite3 database does not exist, a new database named myfriend.sqlite3 is created and SQLite immediately connects to it.

Because you are now connected to the database, you can create tables and other objects in it. If you do not create anything, the database file is not actually created, so create a simple table here. The detailed method for creating tables is explained on another page.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> create table customer (id, name);
$

To exit the command-line tool, enter .exit. This disconnects you from the connected database.

.exit

Now check it in practice. First, connect to the myfriend.sqlite3 database and enter .exit.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>.exit
$

Next, check the database file that sqlite3 created.

$ ls
myfriend.sqlite3 sample.sqlite3   sqldiff          sqlite3          sqlite3_analyzer

A file with the same name as the specified database name has been created. In SQLite, when you create a database this way, a file with the same name is created and all data contained in the database is stored in that file.

If you want to create the database file somewhere other than the directory that contains sqlite3, specify the path together with the database name as follows.

sqlite3 path/database-name

For example, to create a sample.sqlite3 database in the absolute path /Users/devkuma/sqlite/test, run the following command.

$ sqlite3 /Users/devkuma/sqlite/test/sample.sqlite3

About Database Names

You can specify any name you want for a database. In the example above, myfriend.sqlite3 was used, but any format is acceptable.

myfriend
myfriend.db
myfriend.sqlite
myfriend.sqlite3

When you create a database, SQLite creates a file with the same name as the database and stores the data there. For example, if the database name is myfriend, the created file is also named myfriend without an extension, which can make it hard to tell what kind of file it is later.

For that reason, it is a good idea to choose the database name with the stored file name in mind. There are several extensions in common use, but this site uses the .sqlite3 extension consistently.

Connecting to a Database

To perform various operations on a database, you must first connect to it. When you create a new database, SQLite connects to it at the same time. To connect to an already created database, enter the following command at the command prompt.

sqlite3 database-name

As you can see, the format is the same as when creating a new database. If you specify the name of an existing database as the argument, SQLite connects to that database instead of creating a new one.

Now try connecting to the myfriend.sqlite3 database created earlier. Start a command prompt, move to the directory that contains sqlite3, and run the following command.

$ sqlite3 myfriend.sqlite3

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>

To confirm that a new database was not created and that you are connected to the existing database, run the .tables command to check the tables created in the existing database.

.tables

Now check it in practice. First, connect to the myfriend.sqlite3 database and enter .tables.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .tables
customer
sqlite>

You can confirm that the customer table exists. Creating a new database and connecting to an existing database use the same command and the same format, so keep this behavior in mind.

SQLite | Database | Checking Connected Databases

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains how to display information about the currently connected database in SQLite and the file where that database is stored.

Information About Connected Databases

When you are connected to a database, use the SQLite command .databases to display the current database name and the file name where the database is stored.

.databases

Now check it in practice. First, connect to the myfriend.sqlite3 database and enter .databases.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .databases
main: /Users/devkuma/sqlite/myfriend.sqlite3
sqlite>

The following information is displayed on the screen.

main is the database name, and /Users/devkuma/sqlite/myfriend.sqlite3 is the file name where the database is stored.

The term database name may be confusing here. main is the name used when you need to specify a database name in an SQL statement, in the same way that you specify table names and column names. A database connected with sqlite3 database-name is automatically assigned the database name main. In other words, the entry whose database name is main is the connected database.

The name main may seem important now, but you can ignore it for the moment. The details are not covered here, but when attached databases exist, they are also displayed in the list when you run the .databases command. The displayed database name is the name specified when the database was attached.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> 
sqlite> attach database 'myaddress.sqlite3' as address;
sqlite> 
main: /Users/devkuma/sqlite/myfriend.sqlite3
address: /Users/devkuma/sqlite/myaddress.sqlite3
sqlite>

In this way, you can use the .databases command to check information about the currently connected databases.

SQLite | Database | Backing Up and Deleting a Database

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

This page explains how to back up a database created with SQLite by copying the database file, and how to delete a database.

Backing Up a Database

SQLite creates one file for each database, and all information is stored in that file. Therefore, to back up a specific database, you can simply copy the file used by that database.

Now check it in practice. Suppose the myfriend.sqlite3 database has already been created, and connect to this database.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite>

Run the .tables command to check the tables created in the database. A table named customer has been created.

$ sqlite3 myfriend.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .tables
customer
sqlite>

Now disconnect from the database. Then copy the myfriend.sqlite3 file and save it under another name, myfriendbackup.sqlite3. The file name can be anything.

$ cp myfriend.sqlite3 myfriendbackup.sqlite3

This creates a database identical to the original database. Next, connect to the myfriendbackup.sqlite3 database created as the backup and run the .tables command to display the list of tables created in the database.

$ sqlite3 myfriendbackup.sqlite3
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .tables
customer
sqlite>

You can confirm that the same table as in the original database exists. This was only a simple check, but it shows that the database has been duplicated. In SQLite, you can back up a database simply by copying the file that contains the database data.

In addition to this method, SQLite also provides backup and recovery using the .backup and .restore commands.

Deleting a Database

SQLite manages each database independently as a single file. To delete a database, delete the file that stores the database. There is nothing else you need to do in SQLite.

SQLite | Database | Reclaiming Free Space (VACUUM)

kc@example.com (kc kim) — Mon, 14 Oct 2019 23:19:51 +0900

When data is repeatedly added to and deleted from a SQLite database, the file size can become large compared with the amount of data stored. This page explains how to reclaim unused space by using the VACUUM statement.

Using VACUUM

When you create a SQLite database, one file is created, and tables and stored data are saved in that file. As you add data to tables, the database file gradually grows. However, even if you delete data from a table, the database file size does not immediately become smaller. This is because the space used in the file is not removed immediately, but is kept so it can be reused when new data is added.

Normally you do not need to worry about this, but if the file size is a concern, you can run the VACUUM statement to reclaim unused space and reduce the file size.

Run the VACUUM statement as follows.

VACUUM;

When you run the VACUUM statement, SQLite temporarily moves the database contents to a temporary database and then moves them back. This removes free space and rewrites the data sequentially.

One point to note is that VACUUM applies only to the main database. Attached databases are not processed by VACUUM. Also, for tables that do not have a column set as INTEGER PRIMARY KEY, the ROWID assigned to stored data may change.

Practice

Now try it in practice. Suppose there is a sample.sqlite3 file whose database file size is 16,384 bytes.

$ ls -al
total 5048
drwxr-xr-x@ 7 kimkc  staff      238 10 19 23:20 .
drwxr-xr-x  9 kimkc  staff      306 10 17 23:49 ..
drwxr-xr-x  6 kimkc  staff      204 10 19 23:20 bak
-rw-r--r--  1 kimkc  staff    16384 10 19 23:20 sample.sqlite3
-rwxr-xr-x@ 1 kimkc  staff   691768 10 11 18:31 sqldiff
-rwxr-xr-x@ 1 kimkc  staff  1152260 10 11 18:32 sqlite3
-rwxr-xr-x@ 1 kimkc  staff   719796 10 11 18:31 sqlite3_analyzer

Connect to the database and delete one of the tables in it.

$ sqlite3 sample.sqlite3 
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> .tables
select    user      username
sqlite> drop table user;
sqlite>

After deleting the table, check the database file size. It is still 16,384 bytes, the same as before the deletion.

$ ls -al
total 5048
drwxr-xr-x@ 7 kimkc  staff      238 10 19 23:24 .
drwxr-xr-x  9 kimkc  staff      306 10 17 23:49 ..
drwxr-xr-x  6 kimkc  staff      204 10 19 23:20 bak
-rw-r--r--  1 kimkc  staff    16384 10 19 23:24 sample.sqlite3
-rwxr-xr-x@ 1 kimkc  staff   691768 10 11 18:31 sqldiff
-rwxr-xr-x@ 1 kimkc  staff  1152260 10 11 18:32 sqlite3
-rwxr-xr-x@ 1 kimkc  staff   719796 10 11 18:31 sqlite3_analyzer

In this way, deleting a table or data does not immediately reduce the database file size. Now connect to the database again and run the VACUUM statement.

$ sqlite3 sample.sqlite3 
SQLite version 3.19.3 2017-06-27 16:48:08
Enter ".help" for usage hints.
sqlite> vacuum;
sqlite>

After running the VACUUM statement, check the database file size. You can see that the file size has been reduced to 12,288 bytes.

$ ls -al
total 5040
drwxr-xr-x@ 7 kimkc  staff      238 10 19 23:29 .
drwxr-xr-x  9 kimkc  staff      306 10 17 23:49 ..
drwxr-xr-x  6 kimkc  staff      204 10 19 23:20 bak
-rw-r--r--  1 kimkc  staff    12288 10 19 23:29 sample.sqlite3
-rwxr-xr-x@ 1 kimkc  staff   691768 10 11 18:31 sqldiff
-rwxr-xr-x@ 1 kimkc  staff  1152260 10 11 18:32 sqlite3
-rwxr-xr-x@ 1 kimkc  staff   719796 10 11 18:31 sqlite3_analyzer
kimkcui-MacBook-Pro:sqlite-tools-osx-x86-3300100 kimkc$

Running VACUUM reclaims unused space and reduces the file size. If the file size is a concern, run VACUUM.