Compare Tables with capability to sort columns within result

kwsmithkwsmith Posts: 6
edited February 2, 2009 2:50PM in SQL Compare Previous Versions
I have quite a few tables with a large number of columns. Version 7.x is good with showing the results of the comparison based upon column order. I'd like an option that allows me to change the output to sort the columns of both tables so I can more easily match columns in both tables even though they were added in a different sequence. I currently have to copy the columns to word, sort by paragraph, save to a text file for both sides and then use beyond compare to more easily match columns and then make sure the data types and other attributes match. If this capability already exists then I didn't catch it in the interface.

Thanks,

Kevin

PS: a similar sort of capability within the constraints and index output windows where the names of the indexes are ignored and a match is done based upon the table and column names. In other words. Indexes that do the same thing in both tables, but the index names are different. Gaps on the other side would only be for when index columns are different.

Comments

  • SQL Compare, by default, will not parse different column order of tables as having a real difference i.e. they will be seen as identical objects in the top pane and the column order won't be changed when synchronizing. However, if you select the option to 'Force column order' they will be shown as different objects in the top pane of the results and this column order will be synchronized between the two tables.

    The bottom pane (SQL Differences) is purely a textual comparison. It will show you the differences even if they aren't seen as real differences in the top pane. Would you like, in some way, to be able to rearrange and sort the text in the SQL Differences pane so that it matches up the columns and makes it easier to see the differences?
  • I would like, to be able to sort the contents of the SQL Difference pane for indexes and constraints so that the name is ignored and the sort is based upon the statement itself, including table and columns. Then a match up is done between the two panes. If names are the same (everything else is same too) then there is a match and no highlight on either side. If the names are different then the names are highlighted. If the statement is missing in one table, then there is a gap across the statement in the other table. This is selectable option A. Selectable option B would be to try to match on the constraint/key names. If the names on both side exists, the statements would be matched up, even if the columns are different with the differences highlighted. The third selectable options is to compare names that are substrings of the other. For Example. ConstraintA in one table and XConstraintA in a second table. Since ConstraintA is a substring of XConstraintA, both statements would be matched and differences, if any, highlighted. I just compared 200+ tables in different databases and had to manually reconcile each of these three scenarios to determine the real differences between tables on a test system and tables that are the release standard.
  • This is currently not possible in SQL Compare, so I've created a feature request for you. The developers will consider this when working on the next version of SQL Compare. For your reference the feature tracking code is: SC-4296
Sign In or Register to comment.