sink - A data access layer handling synchronization, caching and indexing

	Commit message (Collapse)	Author	Age
...
*	Moved query logic to EntityReader to make it reusable in the resource.	Christian Mollekopf	2016-06-09
\|
*	These checks are no longer necessary	Christian Mollekopf	2016-06-05
\|
*	Make different query runners distinguishable	Christian Mollekopf	2016-05-31
\|
*	Detect and recover from invalid database environment.	Christian Mollekopf	2016-05-30
\| \| \| \| \| \|	Sometimes wrong databases are returned for the name, probably related to threading/incorrect usage of lmdb. For the time being we recover from that by detecting it and retrying.
*	Don't report duplicates in a full scan.	Christian Mollekopf	2016-05-20
\|
*	Print the key with the error message	Christian Mollekopf	2016-05-09
\|
*	Properly count replayed results.	Christian Mollekopf	2016-05-06
\| \| \| \|	We used to count one too many.
*	Account filter for resources and contains comparator in query	Christian Mollekopf	2016-05-01
\|
*	The offset is per parent	Christian Mollekopf	2016-04-29
\|
*	Increase the offset by the actually replayed items.	Christian Mollekopf	2016-04-29
\|
*	Ensure we get something reasonable for uid and type	Christian Mollekopf	2016-04-17
\|
*	Fromatted the whole codebase with clang-format.	Christian Mollekopf	2016-03-03
\| \| \| \|	clang-format -i /*{.cpp,.h}
*	Fetch more data on demand	Christian Mollekopf	2016-02-20
\| \| \| \| \| \| \| \|	We skip values we've already seen and only retrieve the new ones. This currently only properly works in a non-live query and we don't give the model any feedback when we can't fetch more data anymore. However, it generally works and we get the desired effect.
*	Cleanup	Christian Mollekopf	2016-02-20
\|
*	Actually avoid superfluous readEntity calls.	Christian Mollekopf	2016-02-19
\| \| \| \| \|	We still read all values, but just reported the ones before the limit. With this we query 1000 out of 50k values in 63ms.
*	Don't assert in the tests	Christian Mollekopf	2016-02-17
\|
*	Prepared sort indexes	Christian Mollekopf	2016-02-16
\|
*	Ensure we correctly sort by date.	Christian Mollekopf	2016-02-15
\|
*	Prepared querying of batches and added a switch to enable/disable	Christian Mollekopf	2016-02-15
\| \| \| \|	sorting
*	Cleanup	Christian Mollekopf	2016-02-13
\|
*	Prepare query sorting	Christian Mollekopf	2016-02-12
\|
*	More queryrunner timings	Christian Mollekopf	2016-02-11
\|
*	Adapt to latest kasync version	Christian Mollekopf	2016-02-10
\|
*	Support result property transformations	Christian Mollekopf	2016-02-05
\| \| \| \| \| \|	This can be used to modify each result before reporting it to the client. Alternatively this could also be done in the DomainTypeAdaptor, which would perhaps be the cleaner solution...
*	Use default captures where it makes sense	Christian Mollekopf	2016-02-05
\|
*	Introduced debug areas and prettified output.	Christian Mollekopf	2016-02-04
\|
*	Centralized main database name	Christian Mollekopf	2016-02-01
\|
*	Renamed Akonadi2 to Sink	Christian Mollekopf	2016-01-20
\| \| \| \|	(except for documentation).
*	Use KAsync::Job as abstraction for the threading implementation	Christian Mollekopf	2015-12-27
\|
*	Threaded query runner implementation	Christian Mollekopf	2015-12-27
\| \| \| \| \| \| \| \| \|	All database access is now implemented in threads, to avoid blocking the main thread. The resource communication still resides in the main thread to keep the coordination simple. With it comes a test that ensures we don't block the main thread for too long.
*	Don't crash without metadata.	Christian Mollekopf	2015-12-22
\| \| \| \|	Used in tests.
*	Only load the properties we need.	Christian Mollekopf	2015-12-22
\| \| \| \|	...and adjust the test accordingly with what we expect.
*	The generator should produce values one by one.	Christian Mollekopf	2015-12-20
\| \| \| \| \|	k ...not all at once.
*	Avoid starting resources on non-live queries.	Christian Mollekopf	2015-12-19
\|
*	If we know what we want, we can get directly that.	Christian Mollekopf	2015-12-17
\|
*	Make queries by id work	Christian Mollekopf	2015-12-17
\|
*	Debug output	Christian Mollekopf	2015-12-17
\|
*	Fixed some warnings	Christian Mollekopf	2015-11-30
\|
*	Provide status information about children fetch state	Christian Mollekopf	2015-11-30
\| \| \| \|	The fetch state is per parent.
*	Debug output	Christian Mollekopf	2015-11-30
\|
*	Cleanup	Christian Mollekopf	2015-11-28
\|
*	Introduced a QueryRunner object	Christian Mollekopf	2015-11-27
	The QueryRunner object lives for the duration of the query (so just for the initial query for non-live queries, and for the lifetime of the result model for live queries). It's supposed to handle all the threading internally and decouple the lifetime of the facade.