Algorithm

For each node at the left (resp. the right), the tool will look for the corresponding node to the right (resp. the left). The status for each node can be:

• There is no corresponding node at the right (resp. the left
• The corresponding node is updated
• The corresponding node is identical

General behavior

The rules walk the tree, and for each node at the left, try to find the corresponding node at the right. The node at the right which will be picked is the first node which is a child of the same parent as the node to compare^[1]

See #extension_of_the_searchfor more information

.

Default comparison

A default comparison will be used if no Rules configuration is defined.

Node name

Two nodes which don't have the same node name will be considered as different regardless of the rule.

For example, here no nodes will be found at the right for the node_1 node at the left, regardless of the attributes of the nodes:

root node_1

root node_2

Specificity for the root node

The corresponding root node for the left tree will always be found as the root node at the right tree. If the names of the nodes are different, the state will be considered as updated.

For example the root_1 and root_2 root nodes will be considered as updated:

root_1 node_1

root_2 node_1

Extension of the search

The tool limit the search to children of the parent node in the right tree.

For example, if you have the following left and right nodes (supposing each node has no atributes), the child_2 node will be found because it is the same node in the left tree as in the right tree. As it is exactly the same node, if will be considered identical:

root parent child_1 child_2

root parent child_1 child_2

In the next example, there are two child_1 node at the right, the first one will be picked:

root parent child_1 child_2

root parent child_3 child_1 child_1

In this third example, the child_2 node will be found because it is a child of the same parent node in the right tree, even if it is not in the sam position in the parent node. Depending on the configuration, the node will be considered as identical or updated:

root parent child_1 child_2

root parent child_2 child_1

In the last example, the child_2 node will not be found because it is a child of another parent node in the right tree:

root parent child_1 child_2

root parent child_1 child_3 parent_2 child_2

Children of not found nodes

If the corresponding right node was not found for a parent node in the left tree, all children nodes will be considered as not found too. 5

For example, here parent_1 was not found in the right tree, so all the children of parent_1 will not be found too:

root parent_1 child_1 child_2

root parent_2 child_1 child_2

Rules configuration

The configuration of the comparison is specified in an XML file which defines:

• On which condition two XML nodes can be considered as representing the same element
• On which condition two XML nodes can be considered as updated