1. 20 Sep, 2019 1 commit
  2. 19 Sep, 2019 3 commits
  3. 06 Aug, 2019 1 commit
    • Kai Uwe Broulik's avatar
      Prefer categories with the query string visible in results · aa9e1d6c
      Kai Uwe Broulik authored
      This logic was present in the old model and keeps e.g. Baloo file search results from getting higher priority than likely more useful
      application results with the given name. It only affects the order of categories, e.g. scoring "Applications" higher than "Documents"
      if such a match exists.
      The order of matches within a category is untouched and remains determined by the runner based on type and relevance.
      Since the "relevance" of a QueryMatch is global and determined freely by the runners which don't know what other results
      might already be present, we have to resort to some fuzzy logic like this.
      Differential Revision: https://phabricator.kde.org/D22888
  4. 31 Jul, 2019 1 commit
  5. 29 Jul, 2019 1 commit
  6. 26 Jul, 2019 3 commits
    • Kai Uwe Broulik's avatar
      Make runAutomatically more resilient · 269d9825
      Kai Uwe Broulik authored
      We need to delay running the result or else we might end up clearing the model
      whilst in the middle of processing a model insertion
    • Kai Uwe Broulik's avatar
      Fix warnings · 26e30079
      Kai Uwe Broulik authored
    • Kai Uwe Broulik's avatar
      Fix running results automatically · 4d3c62f5
      Kai Uwe Broulik authored
      runCurrentIndex() relies on the ListView being up to date.
      However, all we end up doing is running the first result, so we might as well
      bypass ListView and call the model directly.
      Also, quit the match session before beginResetModel(), otherwise we might get
      results added while the model is in the process of resetting, causing crashes
  7. 25 Jul, 2019 10 commits
  8. 24 Jul, 2019 3 commits
  9. 21 Jul, 2019 1 commit
    • David Edmundson's avatar
      Fix assert when pressing enter really quickly · 21a7ef11
      David Edmundson authored
      runCurrentIndex checks if "currentItem" is valid.
      currentItem relies on the ListView being up-to-date.
      The ListView also updates in response to QAIM::modelReset, the same
      connection this code is attached to, so run afterwards.
      This means the guard for currentItem is potentially out of date, and we
      will call SourcesModel::run without a valid index on an empty source
      model. This hits an assert.
      modelReset and countChanged are identical as update resets
      the model (which is wrong but a task for another day)
      Test Plan:
      Could reproduce the crash by typing ":q <enter>" really quickly
      Could no longer reproduce
      Reviewers: #plasma
      Subscribers: plasma-devel
      Tags: #plasma
      Differential Revision: https://phabricator.kde.org/D22612
  10. 18 Jun, 2019 4 commits
  11. 10 Jun, 2019 3 commits
  12. 09 Jun, 2019 1 commit
  13. 06 Jun, 2019 1 commit
  14. 16 May, 2019 3 commits
  15. 09 May, 2019 1 commit
  16. 05 Mar, 2019 1 commit
  17. 01 Mar, 2019 1 commit
  18. 27 Feb, 2019 1 commit