Split handling of on-demand extractor process from IndexedDataRetriever

Authored by bruns on Nov 11 2019, 12:54 AM.

Description

Split handling of on-demand extractor process from IndexedDataRetriever

Summary:
To allow a unification of the IndexedDataRetriever job and FileFetchJob,
move the process handling into a separate class and only keep
the UserMetaData handling.

As a small optimization, interleave the forking of the external process,
UserMetaData retrieval and joining of the external process. Also reuse
the QFileInfo in the constructor to determine if the file is writable and
thus has editable metadata.

Beyond the optimization, no user visible changes.

Depends on D25247

Test Plan:
When using a location not indexed by baloo, metadata is
still shown as before

Reviewers: Baloo, ngraham, astippich

Reviewed By: Baloo, ngraham, astippich

Tags: Baloo

Differential Revision: https://phabricator.kde.org/D25248

Details

Committed
brunsNov 13 2019, 11:53 PM
Reviewer
Baloo
Differential Revision
D25248: Split handling of on-demand extractor process from IndexedDataRetriever
Parents
R824:bb533b2d2ddb: Add categorized logging for baloo-widgets
Branches
Unknown
Tags
Unknown