Data Profiler is an open source .NET based command line tool (and library) for profiling delimited files, Excel (the 1st sheet), and database tables. It uses Transformalize and Junk Drawer to do most of it's work.

It is released under GNU General Public License, version 3 (GPL-3.0).

The command line tool is dp.exe. It takes 1 to 3 arguments:

####Argument 1, Input: Input should be a file name or fully qualified database table name.

file:

dp c:\temp\fantasy\fantasy2013.txt

table:

dp localhost.NorthWind.dbo.Customers

The fully qualified database table name must indicate the Server, the Database, the Schema, and the Table name. It must be delimited by dots. For example, Server.Database.Schema.Table is correct. When your server name contains dots, surround it with brackets (e.g. [Server.com]). If your provider doesn't support schema, use Server.Database.Table.

By default, database objects assume a provider of sqlserver. You may configure it to expect mysql, or postgresql as well. If you're not using SQL Server with trusted authentication, you may configure ports, user names, and passwords as needed in dp.exe.config.

####Argument 2, Output:

Output is optional. If you don't provide it, it will be whatever is defined as output in your DataProfiler process (see dp.exe.config). It is usually data-profile.html. If your output file has an .html extension, it is rendered and opened as an HTML table with Bootstrap styling. You may also choose a .txt or .csv extension and it will open in NotePad or Excel respectively.

####Argument 3, Sample Percentage:

Sample Percentage is optional. If you don't need a perfect data profile, use an integer between 1 and 99 to sample a percentage of the data. This can really speed things up on bigger files (at the cost of accuracy though).

####dp.exe.config

<transformalize>
	<processes>
		<!-- JunkDrawer process is for profiling files -->
		<add name="JunkDrawer">
			<connections>
				<add name="input" provider="file" file="*"/>
				<add name="output" provider="internal"/>
			</connections>
		</add>
		<add name="DataProfiler">
			<connections>
				<!-- by default, provider is sqlserver -->
				<add name="input" />

				<!-- for other database providers -->
				<!-- <add name="input" provider="mysql" port="*" user="*" password="*" /> -->
				<!-- <add name="input" provider="postgresql" port="*" user="*" password="*" /> -->

				<!-- default output is data-profile.html, in the same folder your dp.exe is in -->
				<add name="output" provider="html" file="data-profile.html" />
			</connections>
			<actions>
				<!-- remove this if you don't want data-profile.html to open after profile is complete -->
				<add action="open" connection="output" />
			</actions>
			<entities>
				<!-- you can control the default sample size here, and whether or not nolock hint is used -->
				<add name="*" sample="100" no-lock="true" />
			</entities>
		</add>
	</processes>
</transformalize>