Improved project file format.
lewismoten
Posts: 58
My goal is to automate documenting all objects under tables, views, programmability, storage, and security within a single database catalog. In other words - document all objects within a single database catalog.
I have a continuouse integration server setup to build my SQL documentation through the command line. The problem that I am having is that each time a new object is added to my database, I need to manually edit the sqldoc file and add the new object. This is counter productive in an agile development environment.
Upon opening the file, I noticed that it was xml, but was not compatable with generally matching all objects within a single database catalog. In addition, the way that the xml is formatted causes problems with viewing the xml file as text. The "Objects" element is 538,567 characters long on a single line for my database.
Here is my suggestion. First, make the xml managable via text readers such as notepad and visual studio. Break the objects tag apart.
<Objects>
<Object name="..." longname="..." type="..." filename="..." fqname="..." owner="..." issystem="..." parentname="..." parenttype="..." extra="..." />
<Object name="..." longname="..." type="..." filename="..." fqname="..." owner="..." issystem="..." parentname="..." parenttype="..." extra="..." />
...
</Objects>
Next - give us some wild cards.
<ObjectMatch name="%" longname="%" type="Function_Scaler" filename="" fqname="%" owner="dbo" issystem="False" parentname="%" parenttype="%" extra="%" />
As I indicated initially, my main goal is to automate a build to have it scan the database for new objects so that they too are included in the documentation. Anything that will do that would be helpful in getting the job done. Output to the log file stating that "WARN: documentation is missing for [object]" would be great for tracking down missing information.
I have a continuouse integration server setup to build my SQL documentation through the command line. The problem that I am having is that each time a new object is added to my database, I need to manually edit the sqldoc file and add the new object. This is counter productive in an agile development environment.
Upon opening the file, I noticed that it was xml, but was not compatable with generally matching all objects within a single database catalog. In addition, the way that the xml is formatted causes problems with viewing the xml file as text. The "Objects" element is 538,567 characters long on a single line for my database.
Here is my suggestion. First, make the xml managable via text readers such as notepad and visual studio. Break the objects tag apart.
<Objects>
<Object name="..." longname="..." type="..." filename="..." fqname="..." owner="..." issystem="..." parentname="..." parenttype="..." extra="..." />
<Object name="..." longname="..." type="..." filename="..." fqname="..." owner="..." issystem="..." parentname="..." parenttype="..." extra="..." />
...
</Objects>
Next - give us some wild cards.
<ObjectMatch name="%" longname="%" type="Function_Scaler" filename="" fqname="%" owner="dbo" issystem="False" parentname="%" parenttype="%" extra="%" />
As I indicated initially, my main goal is to automate a build to have it scan the database for new objects so that they too are included in the documentation. Anything that will do that would be helpful in getting the job done. Output to the log file stating that "WARN: documentation is missing for [object]" would be great for tracking down missing information.
Comments
As far as I'm aware, new objects are select for documentation by default. Is this not the case for you?
Could you please post the command line parameters that you are using?
Thanks,
David Atkinson
Red Gate Software
Product Manager
Redgate Software