NAME

node-odbc - An asynchronous Node interface to unixodbc and its supported drivers

SYNOPSIS

var util = require('util');
var odbc = require("odbc");

var db = new odbc.Database();
db.open("DRIVER={FreeTDS};SERVER=host;UID=user;PWD=password;DATABASE=dbname", function(err)
{
	db.query("select * from table", function(err, rows, moreResultSets)
	{
		util.debug(util.inspect(rows));
		db.close(function(){});
	});
});

DESCRIPTION

unixODBC binding to node. Needs a properly configured odbc(inst).ini. Tested locally using the FreeTDS and Postgres drivers.

INSTALLATION

• Make sure you have the unixODBC binaries and unixODBC headers installed and the drivers configured.
• On ubuntu and probably most linux distros the unixODBC header files are in the unixodbc-dev package (apt-get install unixodbc-dev)
• On OSX one can use macports.org to install unixODBC (sudo port unixODBC)

###git

git clone git://github.com/w1nk/node-odbc.git
cd node-odbc
node-waf configure build

###npm

npm install odbc

TIPS

• If you are using the FreeTDS ODBC driver and you have column names longer than 30 characters, you should add "TDS_Version=7.0" to your connection string to retrive the full column name.

###Example

"DRIVER={FreeTDS};SERVER=host;UID=user;PWD=password;DATABASE=dbname;TDS_Version=7.0"

BUGS

None known, but there might be one ;).

COMPLETE

• Connection Management
• Querying
• Database Descriptions
• Binding Parameters (thanks to @gurzgri)

TODO

• Option to emit on each record to avoid collecting the entire dataset first and increasing memory usage
• More error handling.
• Tests
• SQLGetData needs to support retrieving multiple chunks and concatenation in the case of large column values

ACKNOWLEDGEMENTS

• orlandov's node-sqlite binding was the framework I used to figure out using eio's thread pool to handle blocking calls since non blocking odbc doesn't seem to appear until 3.8.

AUTHORS

Lee Smith (notwink@gmail.com)

Dan VerWeire (dverweire@gmail.com)