There’s a few exciting blog posts coming up including "The AdventureWorks of Tweet-SQL" and "Twitter bots with Tweet-SQL". In the meantime checkout the following blog posts…

MULTIPLE TWITTER ACCOUNTS WITH TWEET-SQL

CONFIGURING OAUTH FOR TWEET-SQL

CREATING A SQLSERVERPEDIA LIST WITH TWEET-SQL

Post from: youdidwhatwithtsql.com

Tweet-SQL 3.5 Released

Firstly, assuming you’ve already gone through the oAuth process for Tweet-SQL once we need to copy the current active account. Just run the below TSQL in SSMS against the database you have installed the software in.

EXEC dbo.tweet_cfg_copy_oauth_account;

This will copy the Twitter account details from the registry to a table called dbo.tweet_oauth_accounts. This account remains active for Tweet-SQL.

Next we need to authorise the additional account with the Tweet-SQL configuration program. Launch the tool, change the username field to the new account screen name, click File > Save Config.

Click oAuth > Get Access Token to begin the oAuth process. The following dialog will appear. Click ‘Yes’. Provided you have executed the dbo.tweet_cfg_copy_oauth_account procedure the old account details will be saved.

In a few moments your browser will launch and you will be prompted to login to Twitter. Login with the appropriate account, authorise Tweet-SQL and copy the pin.

Return to the Tweet-SQL configuration program and paste the pin into the dialog box.

Click ‘OK’ and a success dialog should appear.

Return to SSMS and run the below TSQL to add the new account to the dbo.tweet_oauth_accounts table

EXEC dbo.tweet_cfg_copy_oauth_account;

Now you should have two accounts available for Tweet-SQL.

SELECT twitter_username 
FROM dbo.tweet_oauth_accounts;

To select a particular twitter account we run….

EXEC dbo.tweet_cfg_change_oauth_account 'rhyscampbell';

Then we can post a status from this account;

EXEC dbo.tweet_sts_update 'Sending a test status...', null;

Then to change accounts…

EXEC dbo.tweet_cfg_change_oauth_account 'tweetsql';
EXEC dbo.tweet_sts_update 'Testing the new multiple account support for oAuth in Tweet-SQL.', null;

Change back once more…

EXEC dbo.tweet_cfg_change_oauth_account 'rhyscampbell';
EXEC dbo.tweet_sts_update 'Sending another test status...!', null;

That’s it! There’s no limits to the number of accounts this method will support, 2 or 2, 000! The next version of Tweet-SQL will be released in the coming weeks. In the meantime have some fun with the current version.

Post from: youdidwhatwithtsql.com

Multiple Twitter accounts with Tweet-SQL

Here’s a quick walk-through of how we can get Twitter search results into a database table using Tweet-SQL. First create a table to hold the tweets.

CREATE TABLE dbo.TwitterSearchResults
(
	Id INTEGER NOT NULL IDENTITY(1,1) PRIMARY KEY CLUSTERED,
	twitter_id VARCHAR(100),
	published VARCHAR(30),
	title VARCHAR(160),
	content VARCHAR(160),
	updated VARCHAR(30),
	author_name VARCHAR(50),
	uri VARCHAR(100),
	keyword_searched VARCHAR(100)
);

Now we can run the below script to use Tweet-SQL to store twitter search results in the above table. Just change the value set for @keyword.

DECLARE @xml XML,
        @handle INT,
        @keyword VARCHAR(100);
 
-- Set keyword to search for
SET @keyword = 'Obama'
 
-- Turn off resultset from Tweet-SQL
EXEC dbo.tweet_cfg_resultset_send 0;
 
-- Peform a twitter search
EXEC dbo.tweet_src_search @keyword, null, @xml OUTPUT;
 
-- Prepare an xml document
EXEC sp_xml_preparedocument @handle OUTPUT, @xml, '<root xmlns:a="http://www.w3.org/2005/Atom" />';
 
 
INSERT INTO dbo.TwitterSearchResults
(
	twitter_id,
	published,
	title,
	content,
	updated,
	author_name,
	uri,
	keyword_searched
)
SELECT [a:id],
	   [a:published],
	   [a:title],
	   [a:content],
	   [a:updated],
	   [author_name],
	   [uri],
	   @keyword -- Store the keyword associated with this search
FROM OPENXML (@handle, '//a:feed/a:entry', 2) -- Path to tweets in the xml docs
WITH
(
	[a:id] VARCHAR(160),
	[a:published] VARCHAR(160),
	[a:title] VARCHAR(160),
	[a:content] VARCHAR(160),
	[a:updated] VARCHAR(160),
	author_name VARCHAR(100) 'a:author/a:name', -- Query further down the xml doc to get author info
	uri VARCHAR(100) 'a:author/a:uri'
 );
 
-- Clean up            
EXECUTE sp_xml_removedocument @handle;
 
-- Turn on resultset from Tweet-SQL
EXEC dbo.tweet_cfg_resultset_send 1;

Now the TwitterSearchResults table should contain and bunch of tweets.

SELECT * 
FROM dbo.TwitterSearchResults;

Get yourself a copy of Tweet-SQL and try this out now!

Post from: youdidwhatwithtsql.com

Tweet-SQL: Storing searches in a table

• Full 64 bit OS Support. (Previously the installer would fail to created the required registry keys on 64 bit Operating Systems).
• Full support for oAuth authentication.
• Support for the new Twitter List methods. See sqlserverpedia list with Tweet-SQL for this in action.
• Support for geo-tagging of status updates.
• Support for the Twitter reweet methods.
• Support for the saved search API methods.
• Support for the suggested users API methods.
• Support for TinyURL, TweetShrink and TwitterCounter.

I’ll be following up with a few cool blog posts showing what you can do with these new features. In the meantime checkout the Tweet-SQL User Documentation for more details and download Tweet-SQL Version 3 now!

Post from: youdidwhatwithtsql.com

Tweet-SQL Version 3 released

First copy the list of users from the sqlserverpedia page.

Paste this into Excel and you should get something looking like below.

Remove the section titles, empty rows and any text after the url so we are just left with a list of Twitter profile pages.

Next we need to extract the Twitter username from the url. This little bit of Excel wizardry should do it.

=MID(A1,SEARCH("http://twitter.com/",A1)+19, LEN(A1) - 19)

This formula may break as the page changes in the future so watch out for this. Review the list and remove any invalid values. At the time of writing there’s one of the list that doesn’t contain a twitter url. Save this as a csv and then import it into a database containing the Tweet-SQL Procedures. I’ve uploaded a copy of the file I produced today here. This file contains 261 Twitter users. I used the below table structure.

USE [TweetSQLV3]
GO
 
/****** Object:  Table [dbo].[sqlserverpedia]    Script Date: 07/24/2010 16:16:24 ******/
SET ANSI_NULLS ON
GO
 
SET QUOTED_IDENTIFIER ON
GO
 
SET ANSI_PADDING ON
GO
 
CREATE TABLE [dbo].[sqlserverpedia](
	[url] [varchar](255) NULL,
	[tweep] [varchar](255) NULL
) ON [PRIMARY]
 
GO
 
SET ANSI_PADDING OFF
GO

There’s a few dupes in the below list, because people are listed in multiple sections, so run the below TSQL script to de-duplicate it.

ALTER TABLE dbo.sqlserverpedia ADD Id INTEGER IDENTITY(1,1) NOT NULL PRIMARY KEY;
GO
 
DELETE t1
FROM dbo.sqlserverpedia AS t1
INNER JOIN dbo.sqlserverpedia AS t2
	ON t1.tweep = t2.tweep
WHERE t1.id < t2.id;
 
-- Additional column we'll use later
ALTER TABLE dbo.sqlserverpedia ADD done BIT DEFAULT 0;

Next we’ll need to create the list. The below TSQL will do this (for SQL 2008). Make a note of the list id as we’ll need this later.

DECLARE @list_name VARCHAR(30) = 'sqlserverpedia list';
-- Turn on relational resultsets in Tweet-SQL
EXEC dbo.tweet_cfg_resultset_send 1;
-- Create the list
EXEC dbo.tweet_list_post_lists @list_name, 'public', null;

Next we’ll add the users onto this list. After each record is added to the Twitter list it is flagged as done so you can simply re-run the script if something fails halfway through.

DECLARE @tweep VARCHAR(30), @list_id INTEGER;
 
-- Set your list id here
SET @list_id = 17542298;
 
-- Turn off resultsets in Tweet-SQL
EXEC dbo.tweet_cfg_resultset_send 0;
 
DECLARE tweeps CURSOR LOCAL FAST_FORWARD FOR SELECT tweep
					 FROM dbo.sqlserverpedia
					 WHERE done IS NULL;
 
-- Open the cursor and get the first result
OPEN tweeps;
FETCH NEXT FROM tweeps INTO @tweep;
 
WHILE(@@FETCH_STATUS = 0)
BEGIN
 
	-- Add the tweep to the list
	EXEC dbo.tweet_list_post_list_members @list_id, @tweep, null;
	-- Flag the current record as done
	UPDATE dbo.sqlserverpedia
	SET done = 1
	WHERE tweep = @tweep;
	-- Wait for a bit so we don't annoy twitter
	WAITFOR DELAY '00:00:05'
	-- Fetch the next row
	FETCH NEXT FROM tweeps INTO @tweep;
 
END
 
-- Clean up
DEALLOCATE tweeps;
-- Turn Tweet-SQL resultsets back on
EXEC dbo.tweet_cfg_resultset_send 1;

See my finished list here and follow me on twitter for more TSQL tomfoolery!

Post from: youdidwhatwithtsql.com

Creating a sqlserverpedia list with Tweet-SQL

After installing Tweet-SQL launch the configuration program by clicking Start > All Programs > Tweet-SQL > Tweet-SQL Config.

Ensure the Authentication drop down is set to OAuth. Enter your username in the box labelled username. Tweet-SQL needs your username for some api calls to work. Click File > Save Config to save this information.

Next click oAuth > Get Access Token. Tweet-SQL will launch your default web browser to contact Twitter. Enter your login details into twitter and click Allow.

Once you have granted Tweet-SQL access you will be provided with a PIN. Copy this PIN to your clipboard.

Return to the Tweet-SQL configuration program and enter the PIN in the dialog box.

Tweet-SQL will then verify the PIN with Twitter to complete the process. After a few moments a confirmation will be displayed.

You are now ready to start using Tweet-SQL!

EXEC dbo.tweet_sts_update 'Dinner on the river tonight!', null;

You can continue using basic authentication with Twitter until August 16th 2010. The new version of Tweet-SQL supporting oAuth will be released within the next few weeks. Find out more about oAuth and download Tweet-SQL now!

Post from: youdidwhatwithtsql.com

Configuring oAuth for Tweet-SQL

OpenDBDiff

This is an open source project hosted on Codeplex that has database compare and, crucially, synchronisation abilities. I’ve used this tool in a few production deployments and have found it useful. I don’t fully trust it yet, it took a while to trust Redgate, but I haven’t encountered any major difficulties yet. I’ve had issues with synchronisation scripts and dependency ordering but nothing I have found too frustrating.

The user interface is basic, but well designed, and it’s easy to get productive straight away.

You can’t stay within the tool to synchronise your databases but, as there is a non-functional "Actions Report" tab, it seems this is on the roadmap.

Open DBDiff can synchronize the following object types.

• Tables (including Table Options like vardecimal, text in row, etc.)
• Columns (including Computed Columns, XML options, Identities, etc.)
• Constraints
• Indexes (and XML Indexes)
• XML Schemas
• Table Types
• User Data Types (UDT)
• CLR Objects (Assemblies, CLR-UDT, CLR-Stored Procedure, CLR-Triggers)
• Triggers (including DDL Triggers)
• Synonyms
• Schemas
• File groups
• Views
• Functions
• Store Procedures
• Partition Functions/Schemes
• Users
• Roles

I did have trouble connecting to different instances of SQL Server on the same machine. This shouldn’t matter much for production use as we’d normally be synchronising different boxes.

Starinix Database Compare

This is a freeware tool that is able to compare MySQL, Access and SQL Server databases and any combinations thereof. Not sure this would be useful in my case but I guess someone might do (I can already tell you your MySQL and SQL Server procedures do not match!).

Key features

• Take snapshots of the structure of the database and review it later.
• Snapshots can be saved anywhere, allowing you to take snapshots of any database for reviewing later.
• Compare two online database connections.
• Compare an online database connection to an off-line snapshot.
• Compare two off-line snapshots.
• Use integrated or user based security.
• Compares: views, constraints, stored procedures, functions, tables and fields.
• Compare Access database to an SQL Server database.
• Compare Access database to a MySQL database.
• Compare Access database to another Access database.
• Compare MySQL Database to an SQL Server Database.
• Compare SQL Server 2000 to SQL Server 2005.

I’ve done some quick testing of this tool and it does seem to be accurate, spotting all of my edits made to tables, procedures and functions. The interface is ok but does need some polish. For example, it would be a nice touch for the "SQL View" scroll bar to scroll both the source and destination views. Identifying differences to large procedures isn’t as easy as it should be but clicking "Script Comparison" launches an external diff tool that does a better job.

This is strictly a compare tool, meaning there are no synchronisation abilities, so you’ll have to do a little leg work to implement any needed changes.

Overall there’s no clear winner for me out of these two free tools. Both products have easy to use GUIs. OpenDBDiff clearly has the edge as a synchronisation tool but Starinix has a few nice touches like the external diff tool and the snapshot feature.

Post from: youdidwhatwithtsql.com

Free Database Sync Tools

DECLARE @username VARCHAR(30);
-- Set Twitter username
SET @username = 'rhyscampbell';
-- Execute the procedure
EXEC dbo.tweet_util_twitterCounter @username, null, null;

Here’s the raw xml that is accessible by using OUTPUT parameters.

<twittercounter>
  <user_id>297689</user_id>
  <user_name>rhyscampbell</user_name>
  <followers_current>255</followers_current>
  <date_updated>2009-06-06</date_updated>
  <url>http://www.youdidwhatwithtsql.com</url>
  <avatar>70800065/n571986135_2362730_9166_normal.jpg</avatar>
  <follow_days>253</follow_days>
  <started_followers>18</started_followers>
  <growth_since>237</growth_since>
  <average_growth>1</average_growth>
  <tomorrow>256</tomorrow>
  <next_month>285</next_month>
  <followers_yesterday>258</followers_yesterday>
  <rank>90877</rank>
  <followers_2w_ago>18</followers_2w_ago>
  <growth_since_2w>237</growth_since_2w>
  <average_growth_2w>17</average_growth_2w>
  <tomorrow_2w>272</tomorrow_2w>
  <next_month_2w>765</next_month_2w>
  <followersperdate>
    <date2009-05-11>227</date2009-05-11>
    <date2009-05-12>227</date2009-05-12>
    <date2009-05-15>230</date2009-05-15>
    <date2009-05-17>230</date2009-05-17>
    <date2009-05-20>232</date2009-05-20>
    <date2009-05-22>239</date2009-05-22>
    <date2009-05-24>241</date2009-05-24>
    <date2009-05-26>249</date2009-05-26>
    <date2009-05-29>248</date2009-05-29>
    <date2009-05-31>250</date2009-05-31>
    <date2009-06-02>248</date2009-06-02>
    <date2009-06-04>258</date2009-06-04>
    <date2009-06-06>255</date2009-06-06>
  </followersperdate>
</twittercounter>

Post from: youdidwhatwithtsql.com

Using TwitterCounter with Tweet-SQL

DECLARE @shrunk VARCHAR(140),
		@shaved INT;
 
EXEC dbo.tweet_util_tweetShrink 'For some strange reason I have developed a tendency to think every new version of Windows is more of a resource pig.', @shrunk OUTPUT, @shaved OUTPUT;
 
-- The shortened Tweet
SELECT @shrunk AS shrunkTweet;
-- How may characters have been shaved off?
SELECT @shaved AS shaved;

A future version of Tweet-SQL will include this functionality along with a few more cool surprises.

Post from: youdidwhatwithtsql.com

Shrinking Tweets with Tweet-SQL

• tweet_sog_followers – Returns the ids of the authenticating users followers or, if a non-null value is supplied for @user_id_or_nick, the ids of the specified user’s followers.
• tweet_sog_friends – Returns the ids of the authenticating users friends or, if a non-null value is supplied for @user_id_or_nick, the ids of the specified user’s friends.
• tweet_sog_followersTable – Returns the same data from tweet_sog_followers but saves it to a table in the SQL Server database Tweet-SQL is hosted in. The table created is called tweet_followersIds. Any existing table with this name is dropped at the start of the procedure execution.
• tweet_sog_friendsTable – Returns the same data from tweet_sog_friends but saves it to a table in the SQL Server database Tweet-SQL is hosted in. The table created is called tweet_friendsIds. Any existing table with this name is dropped at the start of the procedure execution.

Twitter added these new methods to help out application developers that were finding the API limits too restrictive.

Executing the below T-SQL will fetch all the ids of the people you follow;

-- The ids of who you follow
EXEC dbo.tweet_sog_friendsTable null;

A new table will be created in the same database.

SELECT * FROM tweet_friendsIds;

Getting the list of ids for the Twitter users is just as easy;

-- The ids of all your followers
EXEC dbo.tweet_sog_followersTable null;

This data can be found in a table called tweet_followersTable

SELECT * FROM tweet_followersIds;

Now we have this data it’s easy to find out the users we’re not following back.

-- People you're not following back
SELECT *
FROM dbo.tweet_followersIds
EXCEPT
SELECT *
FROM dbo.tweet_friendsIds;

People you follow and who follow you back.

-- Twitter buddies! You guys should have a beer!
SELECT *
FROM dbo.tweet_followersIds
INTERSECT 
SELECT *
FROM dbo.tweet_friendsIds

and people who don’t follow you back…

-- Who's not following you back?
SELECT *
FROM dbo.tweet_friendsIds
EXCEPT
SELECT *
FROM dbo.tweet_followersIds;

If you really must know the names, of the people who aren’t following back, then run this T-SQL script. Just don’t react like Mark Corrigan would.

-- Really need to know the names of the 
-- people who aren't following you back?
 
-- Insert the people who aren't following
-- you into a new table called NoFollows
SELECT *
INTO NoFollows
FROM dbo.tweet_friendsIds
EXCEPT
SELECT *
FROM dbo.tweet_followersIds;
 
-- Add PK
ALTER TABLE NoFollows ADD CONSTRAINT 
	pk_id PRIMARY KEY CLUSTERED
	(
		Id
	)
	ON [PRIMARY];
 
-- Add a column to hold Twitter name & screen name
ALTER TABLE NoFollows ADD name VARCHAR(50), screen_name VARCHAR(50);
 
-- Create a cursor to cycle through the users
DECLARE twitterCursor CURSOR STATIC FOR SELECT Id
					FROM dbo.NoFollows
					WHERE name IS NULL;
 
DECLARE @id INT, @name VARCHAR(50), @screen_name VARCHAR(50),
	@xml XML, @handle INT;
 
-- Turn of resultsets from Tweet-SQL
EXEC dbo.tweet_cfg_resultset_send 0;
 
-- Open the cursor and get the first result
OPEN twitterCursor;
FETCH NEXT FROM twitterCursor INTO @Id;
 
-- Loop!
WHILE (@@FETCH_STATUS = 0)
BEGIN
 
	-- Get the user information
	EXEC dbo.tweet_usr_show @Id, null, @xml OUTPUT;
	-- Prepare an xml document
	EXEC sp_xml_preparedocument @handle OUTPUT, @xml;
 
	SELECT @name = name, 
		   @screen_name = screen_name
	FROM OPENXML (@handle, '/user', 2)
	WITH
	(
		name VARCHAR(50),
		screen_name VARCHAR(50)
	);
 
	-- Update the table with the retrieved details
	UPDATE dbo.NoFollows
	SET name = @name,
	screen_name = @screen_name
	WHERE Id = @id;
 
	-- destroy the xml document
	EXEC sp_xml_removedocument @handle;
 
	-- Get the next row
	FETCH NEXT FROM twitterCursor INTO @Id;
 
	-- Be nice to Twitter Servers!
	WAITFOR DELAY '00:00:05';
 
END -- EOF While Loop
 
-- Clean up!
CLOSE twitterCursor;
DEALLOCATE twitterCursor;
-- Turn resultsets from Tweet-SQL back on as appropriate
EXEC dbo.tweet_cfg_resultset_send 1;

SELECT *
FROM dbo.NoFollows;

Post from: youdidwhatwithtsql.com

Know who your friends are with Tweet-SQL