mod_privacy: honor the 'order' attribute #2530
Conversation
|
No, this is not efficient. The list should be sorted during iq-set. |
|
Then you would need to guarantee that the backend preserves the order. Is there such guarantee, including possible future backends? In any case, my concern is the bug #2529 . If it is fixed in some other way, I'll be perfectly happy. |
|
@zinid , I think that I can make ordering O(N) and still used at application-time. Will that be acceptable? If so, I'll resubmit this PR. (I think pre-sorting is not a good design. Imagine that you keep it in an SQL and someone decides to modify the data directly in the backend. Then sorting will be lost. I actually did that before, because user-side privacy list editor(s?) suck.) |
crosser
force-pushed
the
order-privacy-lists
branch
from
dc9d46d to
6c65265
Compare
|
@zinid , I pushed a new version of the patch that does not use |
crosser
force-pushed
the
order-privacy-lists
branch
3 times, most recently
from
851c569 to
b3dd089
Compare
crosser
force-pushed
the
order-privacy-lists
branch
from
b3dd089 to
0ee8a76
Compare
When checking a packet against the active privacy list, make sure that the matching item with the lowest 'order' attribute is used. Bug-url: processone#2529 Signed-off-by: Eugene Crosser [email protected]
crosser
force-pushed
the
order-privacy-lists
branch
from
0ee8a76 to
15b84c5
Compare
|
Hi @crosser, many thanks for your contribution! In order for us to evaluate and accept your PR, we ask that you sign a contribution license agreement. It's all electronic and will take just minutes. |
p1bot
added
the
cla-missing
|
You did it @crosser! Thank you for signing the ProcessOne Contribution License Agreement. We will have a look at your contribution! |
p1bot
removed
the
cla-missing
|
@prefiks: What do you think about this PR and Issue? |
When checking a packet against the active privacy list, make sure that
the items in the list are sorted according to the 'order' attribute.
Bug-url: #2529
Signed-off-by: Eugene Crosser [email protected]