BugFix - Filter Root

[alperozturk96]

• Tests written, or not not needed

Problem?

When users navigate to the Shared or Favorite files, the back button action does not display the correct file list.

What does this PR?

• Filters OCFileListAdapter according to the selected drawer menu item

How to Test?

1. Mark some files as Favorite or share some files.
2. Select Favorites or Shared Files from the drawer menu.
3. Navigate to a subdirectory.
4. Press the back button.

[tobiasKaminsky]

Hm. This looks for me like a strange workaround/fix.
If you are back at root folder, then Favorite/Shared search should be applied, so that no manual filtering will be done.

Favorites is searching across all folders, so e.g. those are favorited

• /test
• /a/1.pdf

If you first go to "favorites" you will see both.
But if with this PR, you go "back" from "test", you would only see "test" folder, but not file "1.pdf".

[alperozturk96]

Hm. This looks for me like a strange workaround/fix. If you are back at root folder, then Favorite/Shared search should be applied, so that no manual filtering will be done.

Favorites is searching across all folders, so e.g. those are favorited

• /test
• /a/1.pdf

If you first go to "favorites" you will see both. But if with this PR, you go "back" from "test", you would only see "test" folder, but not file "1.pdf".

I guess this is fixing the problem you described?

fix_filter.mp4

[tobiasKaminsky]

This works now ✔

Can you check this:

• have a folder /a/b, where "b" is favorited
• delete database (adb shell pm clear com.nextcloud.client)
• start app
• go to favorites
• click on "b"
  --> endless loading for me

But when I click back, I can directly see content of "/a"

[alperozturk96]

Previously, onBrowseUp() was running on the main thread, which caused lag. This PR adds more I/O operations to the function due to previous file changes based on the filter type. Therefore, we need to run onBrowseUp() on the I/O thread and update the UI on the main thread.

Note: The endless loading issue is not part of this PR. It can also occasionally be observed on the stable or master branch. During my tests, I encountered it only once.

@tobiasKaminsky

f8edcba

[github-actions]

Codacy

Lint

Type	master	PR
Warnings	59	59
Errors	3	3

SpotBugs

Category	Base	New
Bad practice	64	65
Correctness	63	62
Dodgy code	296	296
Experimental	1	1
Internationalization	7	7
Malicious code vulnerability	1	1
Multithreaded correctness	6	6
Performance	53	53
Security	18	18
Total	509	509