Disclaimer
ProfileToSQL is provided as is and is completely unsupported.

Intellectual property rights. This stylesheet software is part of the ProfileToSQL suite of stylesheets developed by TF Informatik/Thomas Frisendal. All intellectual property rights belong to TF Informatik/Thomas Frisendal.
Test version. This version of the software is a beta 1 test version, which has only been tested on limited amounts of test data. The software is delivered as is with no guarantees of correct functionality in all situations. Specifically, but not exclusively, this version does not support multi-column constructs, which may lead to unpredicatable results.
Limitation of liability. TF Informatik/Thomas Frisendal is in no events liable for any special, incidental, indirect or consequential damages (incl. damages for lost profits, lost data, business interruption, personal injury, or loss of privay) related in any way to the use of or inability to use this software, even if TF Informatik/Thomas Frisendal has been advised of the possibility of such damages.

Free tool for automation of SQL Server repeatable data profiling


The easy way to load Microsoft SQL Server data profiles into a database!

Automate your profiling activities by way of building a data quality repository database in eg. SQL Server.
By doing that you can repeat the profiling tasks and you can compare results over time. This is the obvious way to work seriously with data quality on a repeated basis - enabling you to answer questions like "Is the quality of our Product data increasing over time?"
The current version of the Data Profile Task in SQL Server 2008 does indeed provide you with most of what the need in terms of data quality measurements. The profiling results are kept in XML files, which are nicely structured according to a quite complex schema.
However, as a result of that, it is not straightforward to load the results into a relational database. The challenge is the deep hierarchical structure of the XML files.

Using XSLT stylesheets on Data Profiling XML files

I have made a solution to this problem, which I offer you free of charge (and un-supported). Using XSLT stylesheets you can reformat the XML files produced by the Data Profile Task. The resulting, reformatted, files, are loaded 1:1 into simple tables, one for each type of profile:
  • Candidate Keys
  • Column Null Ratios
  • Column Patterns
  • Column Statistics
  • Functional Dependancies
  • Inclusions (between tables)
  • Lengths of column content
  • Value distributions of column content
The loading can easily be accomplished using normal SSIS data flows, which you must set up yourself. The table design of the data profiing repository database is also left to you to decide.
See the practical usage recommandations here.
How to get them
To get access to the zip file with the eight XSLT stylesheets, you must register here.
The stylesheets are still in test, but seem to work quite well. Carlos Fernandez at Dataprix in Catalunya/Spain has tested them and documented them here. Some of Dataprix’s pages are available in Catalan, Spanish, English and Italian.