How to remove initial row in repeating tables (or delete all rows) with rules only
Special thanks: Karl D. Swartzendruber
First of all you have to place your repeating group into the non repeating one. So your field structure should be looking like that:
Then as you can see the simple rule f=cleaningGroup will delete all rows. Now we are getting to the point why it never been implemented before. As soon as you will try to assign any value to a group InfoPath will yell at you with the message: “You must select a field. Groups do not have values and therefore cannot be assigned to by this action.” Let’s prove that at least “cannot be assigned” part of that statement is false.
It’s time for some rule breaking.
Now we will screw up the group (“f” in our case) by giving it a temporary name. Create a field at the same level as your group formerly named “f”. Name that field … guess what? Right, with the name “f”. Now InfoPath UI won’t be objecting when you’ll try to create button rule f=cleaningGroup. Delete field “f”. Rename your group back to its original name: “f”.
The Form to test. Web browser forms OK.
Deleting single row from the repeating field.
Here is form that is doing exactly this. The way it was created (renaiming hack) is similar to the previous form. There are several limitations assosioated with this approach. First is it supports single repeating field not repeating group. Another “feature” you have to consider before you decide to use this form is unchecking checkbox at single row will delete all rows with the same value (might be useful in certain scenarios). Also because of Multiple selection check box this form can be made browser enabled in SP2010 only.