Thanks Oliver.
I think I understand what you're saying about why Channel 4 would not be
included with the OverrideCombiner (if it doesn't exist in
testfile1.xml). I am sorry to hear, though, that attributes are not used
at all for matching with OverrideCombiner. If I'm understanding it
correctly, that means it can't really be used for overriding just
specific list items, as the entire list would have to be represented, in
the same order, in both primary and overriding files, or the wrong list
item might be overridden. (For example, if the primary list had Channels
1, 2, and 3, and the override list had only Channel 2, then it would
attempt to override Channel 1 with the info from Channel 2.) Is my
understanding correct?
Anyway, I'm going to experiment a bit and try to make sure I have a good
understanding of how this works.
Thanks again for the help.
Greg
On 11/13/2016 01:51 PM, Oliver Heger wrote:
Hi Greg,
Am 13.11.2016 um 18:41 schrieb Greg Torrance:
Hi Oliver,
Thanks for the reply. That is helpful!
Hm, it looks as if the examples are partly incorrect. I assume that
the channel with id="4" should not be part of testfile1.xml, but only be
contained in the second file. It should appear in the results of union
and merge combiner. Does that make sense?
I'm not clear why -- if the channel with id="4" is not contained in
testfile1.xml -- why it will not end up in the results from all three
combiners. Why just union and merge and not override, too?
On a higher level, OverrideCombiner detects that both files have a list
of channels. Therefore, it copies the first list and ignores the second
one. So the channels element in the first file overrides the one in the
second file.
As for the attributes, do I understand correctly that the attribute
values for elements are used, essentially, as "key" values to
distinguish the various elements for combining purposes? And if the
values of any attributes of two elements differ, then the elements would
be considered different elements for combining purposes? (Said another
way, if two elements have a set of attributes in common, if the values
for those common attributes are the same, the elements will be
considered to match.) Does that make sense? Is that correct?
This is true only for MergeCombiner. The other combiners do not try to
match nodes in both configurations based on their attribute values; they
take two source nodes at the same position in the hierarchy and produce
a result node that has all the attributes of the two source nodes.
Oliver
Thanks again for the reply.
Greg
On 11/13/2016 12:05 PM, Oliver Heger wrote:
Hi Greg,
Am 13.11.2016 um 15:07 schrieb Greg Torrance:
Hi all,
I am trying to work out the precise differences between
OverrideCombiner, UnionCombiner, and MergeCombiner. However, the
examples on this page
(https://commons.apache.org/proper/commons-configuration/userguide/howto_combinedconfiguration.html)
are confusing to me.
testfile1.xml and testfile2.xml both contain Channels with id="4",
however none of the Combiner Results shows any Channels with id="4".
And the Notes on the right-hand side of the page contain no mention of
Channel 4.
Am I missing something? Are the examples correct?
Hm, it looks as if the examples are partly incorrect. I assume that the
channel with id="4" should not be part of testfile1.xml, but only be
contained in the second file. It should appear in the results of union
and merge combiner. Does that make sense?
What I'm really trying to understand is the impact of attributes in the
way the various combiners choose to combine elements. (Any pointers to
further documentation on this would be appreciated.)
Some more information can be found in the Javadocs of the combiner
classes:
- UnionCombiner states: "Another limitation is the handling of
attributes: Attributes can only have a single value. So if two nodes are
to be combined which both have an attribute with the same name, it is
not possible to construct a proper union attribute. In this case, the
attribute value from the first node is used."
- OverrideCombiner has an addAttributes() method with the following
documentation: "Handles the attributes during a combination process.
First all attributes of the first node are added to the result. Then all
attributes of the second node, which are not contained in the first
node, are also added."
- The merging behavior of MergeCombiner is also described in its class
documentation.
Does this help?
Oliver
Thoughts?
Thanks,
Greg
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@commons.apache.org
For additional commands, e-mail: user-h...@commons.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@commons.apache.org
For additional commands, e-mail: user-h...@commons.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@commons.apache.org
For additional commands, e-mail: user-h...@commons.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@commons.apache.org
For additional commands, e-mail: user-h...@commons.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscr...@commons.apache.org
For additional commands, e-mail: user-h...@commons.apache.org
