I need SQL functionalities for a Java JSE application, but dont need a whole SQL server, with things like listen on a port, connection string or even a standalone process to be runned or configured.
I also would prefer to work with files as storages, so that file path identifies DB data.
So given to the DB API the name of a file, I would need perform SQL with the file as DB storage, supporting tables, search, joins and inserts, without thinking to things like ports, external processes, server installation, ecc..
Without any other configurration action, since any other configurable feature is not needed.
Is there some library, preferrably installable as single .jar, that provides this functionality?
If there is not this library, which file-based DB is the simplest to configure and use within JSE, and which configuration steps are needed to perform a query in the provided DB and deploy it with (working) the java .jar application?
I suggest Derby db.apache.org/derby/
I like H2 Database very much. It compares very well with other database engines.
Sure, Hypersonic SQL or Derby, the database that comes bundled with Java 7, will both fill the bill. SQLite would be a third alternative.
you may want to check this out:
http://www.sqlite.org/
I have used HSQLDB in the past and liked it. Depending on what your needs are, you might also be interested in JoSQL which allows you to do SQL type of queries on java collections.
Related
I couldn't find an answer to this question. How can I export a Java project that makes use of a PostgreSQL database?
I want to use the same database on another computer. Do I need to export the database itself with the project? How can this be done?
What should the connection URL be, so that the database is accessible on another computer?
I'm using JDBC, and I'm on Windows.
Thanks in advance.
EDIT: Wouldn't I also need to dynamically retrieve the username and password on the other computer, instead of using the specific username and password I have on my computer in PostgreSQL?
It really depends on what you want to achieve.
Shared database between hosts
Do you want the application on both computers to use the same database, so that changes made by one are seen on the other? If so, you need to configure each copy of the application to connect to the same database instance on one of the machines. This is usually done by changing the JDBC URL. You'll need to configure PostgreSQL on the machine that'll be the database server so it allows connections from the other hosts, ensure they can talk to each other over TCP/IP, etc.
Fresh DB on each host
Do you want each install to have a separate instance of the database, so changes made on one have no effect on the other, and where each instance starts out with a blank, empty database (or one with only static contents like lookup tables)? If so, you should generally define the database using SQL scripts, then have the application run the SQL scripts when first installed on a machine. If you've defined the database by hand so far, you can use pg_dump to create a SQL script that you can use as the basis for this, but I really advise you to look into tools like Liquibase for schema management instead.
"Fork" current state
Do you want each instance of the application on a machine to have an independent database, so changes made on one have no effect on other instances on other machines, but where the starting state of an install is what was in the database on the other host? If so, you need to dump and reload the database alongside the application, using pg_dump -Fc and pg_restore. You can automate this within your application / build system using tools like ProcessBuilder, or do it manually.
There's no generic, canned way to do this. It'll require you to define an application deployment procedure, maybe produce an installer, etc.
I wonder if it's possible to use the database file from mysql, without having a server running. Just copying the db file and place it somewhere, then use "jdbc:mysql://localhost:3306/table";
and change it to something like "jdbc:mysql://C:/Users/me/Desktop/table";
Will this work or is there a better way?
This is not possible with Mysql. Sqlite is a Serverless database designed for this purpose.
You're probably looking for an embedded database, i.e. as you say a library that is able to access a database on a file without a server.
MySQL has an embeddable version: http://www.mysql.com/oem/
You might want to check also H2: http://www.h2database.com/
Or Apache Derby: http://db.apache.org/derby/
Or HSQLDB: http://hsqldb.org/
Short answer: no.
Long answer: You'd have to basically recreate the MySQL daemon in Java. In particular, JDBC would have to know the structures inside of the file. The files are quite complicated, and this would be quite a pain.
This means you would have to basically write your own code capable of parsing and manipulating MySQL files. This would be a horribly complex task.
Yes, there is a better way. If you want to use MySQL, use MySQL.
Another option would be to use the embedded version of MySQL, or something like SQLite.
My aim is to create a local database that can be read and written to with Java. I have some experience with manipulating a local sqlite database with Python, and with interacting with existing networked databases on Microsoft Azure via VB.Net, but the Java formulation for creating a database is escaping me.
Most sources (like the JDBC Docs) seem to assume that you are accessing a database through a network protocol, or a database hosted on localhost. My desired implementation is to create and store the database in a file (or collection of files), so that it can be stored and accessed locally, without network connectivity (presumable through the "file:" protocol).
The JDBC Tutorial looks like it will be very useful once I am up and running, but is currently beyond my scope, since I don't even have an existing database yet.
Many sources have suggested solutions like H2, MySQL, Derby, or Hypersonic DB. However, I'm loath to install extensions (if that's the right term) for a number of reasons:
This project is initially intended to help me learn my way around Java - widening the scope of the project will dilute my experience with the "base" language and, probably, increase the temptation to engage in "cargo cult programming"
If this project does ever get distributed to other users (admittedly unlikely, but still!), I don't want to force them into installing more than the core of Java.
I simply don't know how to install extensions (add-ons? modules?) in Java - one baby-step at a time!
For similar reasons, installing Microsoft SQL Server would not be productive.
This answer looks close to what I'm aiming for; however, it gives the error:
java.sql.SQLException: No suitable driver found for jdbc:mysql://localhost/?user=root&password=rootpassword
and trying "jdbc:file://targetFile.sql" gives a similar error.
I've seen the term "embedded" database, which I think is a subset of "local database" (i.e. a local database is stored on the same system - an embedded database is a local database that is only used by a single application) - if I've got those definitions wrong, please feel free to correct me!
Most likely, the reason for which you are getting the error, is due to the fact that you are not registering the driver (using reflection...) before actually using it for establishing a connection and so on.
Presumably you will want to do something along the lines of Class.forName("driver")
and then cast that if necessary and then registering it in the DriverManager before calling the getConnection() method.
Here is a very useful link that might help you out in solving the issue:
http://www.kfu.com/~nsayer/Java/dyn-jdbc.html
However, if you really want to use a local database/file you might want to have a look at SQLite, that might be one way to go about it, although I recommend going for the MySQL approach, as it is a lot easier to configure and learn how stuff works with JDBC.
If you are still considering SQLite check this out:
Java and SQLite
I see you need some guidance in importing external .jar files into your code (i.e. 3rd party libraries like the ones you will be using for a JDBC driver). Are you using an IDE (e.g. Eclipse, Netbeans, etc.) or are you writing in a text editor and compiling manually?
A number of embedded pure Java databases appeared recently, which have a really simple interface, usually just java.util.Map, don't involve using JDBC or other SQL artifacts, and store their data in a single file or directory:
Chronicle Map
JetBrains Xodus
MapDB
The main downside is that most of such databases provide only the simples key-value model.
DBC can be used with any database that has a JDBC driver, which isn't necessarily a database in "network mode", it can be used with embedded databases as well.
Here are some Java and embeddable databases:
http://www.h2database.com/html/main.html
http://db.apache.org/derby/
http://hsqldb.org/
Java's JDK does not include any implementation of a database nor drivers to access it. It only provides JDBC as an abstraction to connect to a "database". Is up to you to include all the needed libraries in your code.
If you want to have a self contained code you can simply include the .jar file of a embeddable database in you classpath. That way you can create the instance of the database in your code and minimize the external dependencies.
You can find here a list of java embeddable databases
You can find here an example of how to embed HSQLDB in your code.
I want to make java code that creates a sample database in 3-4 DBMS like mysql, oracle sql, sql server etc installed on any OS - windows, linux distro, Mac OS etc.
How can I make my code:
automatically (or with help from a user) locate the jdbc driver in the computer.
execute a fixed set of sql commands which will work regardless of the DBMS used.
Please suggest how I can do all these things.
EDIT:
This will be a back end kind of app.
I am a little new to JDBC, so I am looking for simple/elementary solutions to begin with.
Will switch to advanced ones later.
Thanks.
Bundle drivers for all the supported databases with your program. Users shouldn't have to deal with JDBC drivers or connection strings. (Provide a UI to edit the latter, which might differ between the databases.)
Use an ORM (like Hibernate); or, if you don't need to populate the database with data, a database migration library (like Flyway)
1, See here... How to use a JDBC driver from an arbitrary location
2, Different DBMSs use similar but not identical syntax. You have at least 3 options:
only use sql commands that are supported by all the DBMSs you're interested in;
sniff the DBMS and modify your SQL statements accordingly;
use a framework that comes with an SQL abstraction layer (e.g. Java Persistence API's JPQL). I suspect this may be too much work for what you're after.
If you want to let users find the JDBC drivers, then you shall deal with ClassLoaders and implementing custom ClassLoaders, which is not a simple thing to do. Or you shall use a Application Server which will handle this for your.
Otherwise you shall have all your JDBC drivers available in your class-path.
By the way JDBC is an adapter for working with most RDBMSes which work with SQL, each database provider made vendor specific customization to their SQL. For example you have sequences in Oracle and auto-numbers in MySQL. Or you can use limit in MySQL queries which is not available in Oracle. The solution to this problem is doing what Hibernate does (having Dialects for handling database vendor specific stuff.)
I would like to create a Java library without any dependency, but it needs to use an embedded SQL database. Is it good practice to package both the binary database file and the driver to access it inside the JAR file? What are some advantages and disadvantages?
HSQLDB
More and more Java applications have their own 'in-memory' database that is launched upon application start. Check out hsqldb for an example.
Most obvious advantage is that you control the database and don't have to take different DB servers and their sql dialects into account.
Go on, it's a widely adopted practice used by Hibernate for example.
For a memory constrained devices:
Then do not use embedded database at all. Try using ThoughtWorks library XStream which serializes/deserializes objects to XML/jSon which can be stored in files. Very effective solution with a small memory foot print.
Your database will be read-only (since you cant repackage the jar at runtime). If that is fine then go nuts :)