Openbravo Issue Tracking System - Retail Modules
View Issue Details
IDProjectCategoryView StatusDate SubmittedLast Update
0035470Retail ModulesStoreServerpublic2017-03-09 16:482017-03-18 11:49
Reportermtaal 
Assigned ToAugustoMauch 
PrioritynormalSeverityminorReproducibilityhave not tried
StatusclosedResolutionfixed 
PlatformOS5OS Version
Product Version 
Target VersionRR17Q2Fixed in Version 
Merge Request Status
Review Assigned Tomtaal
OBNetwork customer
Support ticket
Regression level
Regression date
Regression introduced in release
Regression introduced by commit
Triggers an Emergency PackNo
Summary0035470: Code should not adapt the Map instance it gets from a method, but instead call the originating class to adapt the Map
DescriptionSee the code here [1], it gets a Map instance from the SynchronizationEngineProvider.getChannelsWithInitialLoadInProgress() method and changes the Map instance. This works as the Map instance is the same as being kept internally in the SynchronizationEngineProvider. But it is less nice from a encapsulation point of view to adapt the map outside of its owner (the SynchronizationEngineProvider).

[1]
https://code.openbravo.com/erp/pmods/org.openbravo.replication.symmetricds/rev/5f4547f93c1a9a8ba8add123b198ec084e8ff95e#l4.42 [^]
Steps To ReproduceSee description
Proposed SolutionBetter to either call the setter again (makes no difference in the logic really but shows that the resulting map is set back, ignoring multi-threading issues).

But nicest probably to have a removeChannel method on the SynchronizationEngineProvider to remove a channel from the internal Map.
Additional Information
TagsNo tags attached.
Relationships
blocks feature request 0034835 closed AugustoMauch Use another approach for the Table Reload process 
Attached Files
Issue History
Date ModifiedUsernameFieldChange
2017-03-09 16:48mtaalNew Issue
2017-03-09 16:48mtaalAssigned To => AugustoMauch
2017-03-09 16:48mtaalTriggers an Emergency Pack => No
2017-03-09 16:49mtaalRelationship addedblocks 0034835
2017-03-10 09:14hgbotCheckin
2017-03-10 09:14hgbotNote Added: 0094894
2017-03-10 09:14hgbotStatusnew => resolved
2017-03-10 09:14hgbotResolutionopen => fixed
2017-03-10 09:14hgbotFixed in SCM revision => http://code.openbravo.com/erp/pmods/org.openbravo.replication.symmetricds/rev/7416d6418ced0a003624169ee3c0f7e12f7a14e0 [^]
2017-03-18 11:48mtaalReview Assigned To => mtaal
2017-03-18 11:48mtaalProposed Solution updated
2017-03-18 11:49mtaalNote Added: 0095383
2017-03-18 11:49mtaalStatusresolved => closed

Notes
(0094894)
hgbot   
2017-03-10 09:14   
Repository: erp/pmods/org.openbravo.replication.symmetricds
Changeset: 7416d6418ced0a003624169ee3c0f7e12f7a14e0
Author: Augusto Mauch <augusto.mauch <at> openbravo.com>
Date: Fri Mar 10 09:14:09 2017 +0100
URL: http://code.openbravo.com/erp/pmods/org.openbravo.replication.symmetricds/rev/7416d6418ced0a003624169ee3c0f7e12f7a14e0 [^]

Fixes issue 35470: Improves encapsulation in server Initial Data Load process

The SynchronizationEngineProvider class keeps a Map to manage the channels whose initial data load is not finished. Before this changeset, this Map was shared with other classes, that r
ead it and modified. Now the map is not shared, and some methods to modify the map have been added to SynchronizationEngineProvider to improve the encapsulation.

---
M src/org/openbravo/replication/symmetricds/SynchronizationEngineProvider.java
M src/org/openbravo/replication/symmetricds/extensions/BatchDatabaseWriterControl.java
---
(0095383)
mtaal   
2017-03-18 11:49   
Reviewed and tested