Method and system for implementing a filter in a data synchronization system

A method for a sync engine to synchronize record data between a dataset, which is managed by a database manager, and a data subset, wherein a filter specifies one or more criteria under which selected records in the dataset are excluded from synchronization with the data subset, the method comprising:

• receiving changes from the data subset and propagating the received changes into the dataset;
• applying the filter criteria to the changes received from the data subset and, for changes that fail the filter criteria, marking the corresponding records in the dataset as being fresh relative to the data subset;
• requesting from the database manager, changes that are fresh relative to the data subset;
• applying the filter criteria to the changes received from the database manager and modifying the changes and the dataset records as required to implement the filter criteria;
• and
• sending changes that survive application of the filter criteria to the data subset.

A method of synchronizing a dataset with a data subset, wherein a changeable filter identifies selected records from the dataset that are to be synchronized with the data subset, the method comprising:

• occasionally synchronizing the selected records from the dataset with the data subset by performing the following substeps:
• (a) identifying one or more records in the dataset that have been modified since a most recent prior synchronization with the data subset as fresh changes;
• (b) applying the current filter to the fresh changes;
• and
• (c) sending the filtered, fresh changes that survive the application of the filter to the data subset;
• and
• when the filter is changed from a previous filter to a new filter, marking records that have not been modified since the most recent prior synchronization with the data subset, that were excluded from the data subset during the most recent prior synchronization and that pass the new filter as having been added to the dataset since the most recent prior synchronization, and marking records that have not been modified since the most recent prior synchronization with the data subset, that were not excluded from the data subset during the most recent prior synchronization and that fail the new filter as having been changed since the most recent prior synchronization.

A method for synchronizing record data from a dataset to a data subset, wherein a filter specifies one or more criteria under which selected records in the dataset are excluded from synchronization with the data subset, the dataset and the data subset having already been synchronized according to the filter criteria during a previous synchronization, the method comprising:

• (a) identifying one or more outgoing changes from one or more dataset records that have been added, modified or deleted since the previous synchronization;
• (b) applying the filter criteria to the outgoing changes and:
• (i) if the outgoing change is an add and the dataset record fails the filter criteria, canceling the add, marking the dataset record as excluded, and marking the record as no longer needing to be synchronized with the data subset;
• (ii) if the outgoing change is a delete and the dataset record was excluded from the data subset during the previous synchronization, canceling the delete;
• (iii) if the outgoing change is an update, the dataset record was excluded from the data subset during the previous synchronization and the dataset record passes the filter criteria, changing the update to an add;
• (iv) if the outgoing change is an update, the dataset record was excluded from the data subset during the previous synchronization and the dataset record fails the filter criteria, canceling the update and marking the dataset record as no longer needing to be synchronized with the data subset;
• or
• (v) if the outgoing change is an update, the dataset record was included in the data subset during the previous synchronization and the dataset record fails the filter criteria, changing the update to a delete;
• (c) sending the outgoing changes that pass filtering to the data subset;
• and
• (d) for each outgoing change that is a delete that was caused by the dataset record failing the filter criteria, marking the dataset record as being excluded from the data subset.

A method for accommodating a change in a data filter in a synchronization system, the synchronization system comprising a sync engine, a dataset and a data subset, the dataset comprising a first plurality of records and the data subset comprising a second plurality of records, the second plurality of records comprising a subset of the first plurality of records, the filter indicating which records in the first plurality of records are to be synchronized with the data subset, wherein one or more records in the first plurality that have not changed since a previous synchronization between the dataset and the data subset are currently in a state such that the records would not be considered for synchronization during a subsequent synchronization, the method comprising the steps of:

• receiving a change in the data filter, determining a first set of one or more records from the first plurality that failed the filter before the filter was changed and so were not synchronized during the previous synchronization and that pass the filter after the filter was changed, and marking the first set of records as newly added records so that the first set of records will be considered for synchronization with the data subset during the subsequent synchronization;
• determining a second set of one or more records from the first plurality that passed the filter before the filter was changed and were synchronized during the previous synchronization and that fail the filter after the filter was changed, and marking the second set of records as having changed since the previous synchronization so that the second set of records will be considered for synchronization with the data subset during the subsequent synchronization;
• and
• performing the subsequent synchronization between the dataset and the data subset according to the changed filter.

A method of synchronizing a dataset with a first data subset and a second data subset, wherein a first filter indicates a first set of records from the dataset that are to be synchronized with the first data subset and a second filter indicates a second set of records from the dataset that are to be synchronized with the second data subset, the method comprising:

• receiving a first set of incoming changes from the first data subset, performing conflict resolution on the first set of incoming changes, applying the first filter to the first set of incoming changes and propagating one or more changes from the first set of incoming changes that survive conflict resolution and filtering into the dataset;
• receiving a second set of incoming changes from the second data subset, performing conflict resolution on the second set of incoming changes, applying the second filter to the second set of incoming changes and propagating one or more changes from the second set of incoming changes that survive conflict resolution and filtering into the dataset;
• determining a first set of outgoing changes to be sent to the first data subset, applying the first filter to the first set of outgoing changes and sending one or more changes from the first set of outgoing changes that survive filtering to the first data subset;
• and
• determining a second set of outgoing changes to be sent to the second data subset, applying the second filter to the second set of outgoing changes and sending one or more changes from the second set of outgoing changes that survive filtering to the second data subset.