Louvorg/ReaderForSelfoss-multiplatform
3
2
Fork 2
Code Issues 7 Pull Requests 2 Actions Releases 84 Wiki Activity

Can not add sources if url does not end with a slash #184

New Issue
Closed
opened 2025-02-28 18:45:38 +00:00 by davidoskky · 8 comments
davidoskky commented 2025-02-28 18:45:38 +00:00
Contributor
Copy Link
Copy Source

Description

If you log in entering the url without a final slash, the login correctly works. However, the url is considered invalid when attempting to add sources.

Steps to Reproduce

  1. Log in introducing the url of the server without final slash
  2. Open the sources page
  3. Tap the + button

Expected behavior: You should be able to add sources anyway.

### Description If you log in entering the url without a final slash, the login correctly works. However, the url is considered invalid when attempting to add sources. ### Steps to Reproduce 1. Log in introducing the url of the server without final slash 2. Open the sources page 3. Tap the + button **Expected behavior:** You should be able to add sources anyway.
AmineB commented 2025-03-02 12:26:25 +00:00
Owner
Copy Link
Copy Source

I couldn't reproduce. Are you sure the issue isn't with the selfoss api ?

I couldn't reproduce. Are you sure the issue isn't with the selfoss api ?
AmineB added the Status = Need more details label 2025-03-02 12:26:43 +00:00
davidoskky commented 2025-03-03 17:34:03 +00:00
Author
Contributor
Copy Link
Copy Source

Strange, I was able to reproduce this on my physical device and on emulated devices with API 25, 34 and 35.
I'm running Selfoss 2.19 stable, but I reckon I do not have a second instance running to test this on. No API call appears to take place, thus I'd exclude it is a problem on server side.

Looking here it appears that urls without a final slash should not even be allowed to log in at all:

ReaderForSelfoss-multiplatform/androidApp/src/main/java/bou/amine/apps/readerforselfossv2/android/utils/LinksUtils.kt Lines 48 to 57 in f38936f9b4
fun String.isBaseUrlInvalid(): Boolean {
val baseUrl = this.toHttpUrlOrNull()
var existsAndEndsWithSlash = false
if (baseUrl != null) {
val pathSegments = baseUrl.pathSegments
existsAndEndsWithSlash = "" == pathSegments[pathSegments.size - 1]
}
return !(Patterns.WEB_URL.matcher(this).matches() && existsAndEndsWithSlash)
}

However, I am able to log in. When attempting to add a source I receive the error message "Log in to add sources." in a toast. The client has correctly done the login since I am able to use the application normally and my selfoss instance does not allow guest access.

Strange, I was able to reproduce this on my physical device and on emulated devices with API 25, 34 and 35. I'm running Selfoss 2.19 stable, but I reckon I do not have a second instance running to test this on. No API call appears to take place, thus I'd exclude it is a problem on server side. Looking here it appears that urls without a final slash should not even be allowed to log in at all: https://gitea.amine-bouabdallaoui.fr/Louvorg/ReaderForSelfoss-multiplatform/src/commit/f38936f9b43c099785c32ff83d5bbdfd64a78f33/androidApp/src/main/java/bou/amine/apps/readerforselfossv2/android/utils/LinksUtils.kt#L48-L57 However, I am able to log in. When attempting to add a source I receive the error message "Log in to add sources." in a toast. The client has correctly done the login since I am able to use the application normally and my selfoss instance does not allow guest access.
AmineB commented 2025-03-03 21:20:23 +00:00
Owner
Copy Link
Copy Source

this.toHttpUrlOrNull() returns the url with a slash.

Can you provide a screengrab of the error, please?

`this.toHttpUrlOrNull()` returns the url with a slash. Can you provide a screengrab of the error, please?
davidoskky commented 2025-03-03 21:25:36 +00:00
Author
Contributor
Copy Link
Copy Source

Here you go, right after tapping the button to add a source I am redirected to the home and the toast appears.

Here you go, right after tapping the button to add a source I am redirected to the home and the toast appears.
Screenshot_20250303_222448.png
69 KiB
Screenshot_20250303_222448.png
AmineB commented 2025-03-06 21:07:33 +00:00
Owner
Copy Link
Copy Source

The isBaseUrlInvalid is called on the login screen, so the result should be the same on the sources screen too.

This isn't an issue with the url.

The `isBaseUrlInvalid` is called on the login screen, so the result should be the same on the sources screen too. This isn't an issue with the url.
AmineB commented 2025-03-06 21:08:08 +00:00
Owner
Copy Link
Copy Source

I'll close for now. Please reopen if you have more details.

I'll close for now. Please reopen if you have more details.
davidoskky commented 2025-03-09 12:27:00 +00:00
Author
Contributor
Copy Link
Copy Source

So, I went through both checks with the debugger.
isBaseUrlInvalid does return true in both cases when you use a url without a final slash. However in the login process, that does not stop the login attempt. Login is successful and thus you proceed to the home page. In the source adding page, if the base url is considered invalid then it immediately kicks you out without attempting to perform any operation.

It is supposedly a bit strange, because this is the only place in the application where we check the validity of the url after we already did log in. I'm not really sure the check itself is necessary and may be removed altogether from the sources. I think the check is meant to ensure we actually did log in and are not connected as a guest, but that is not really what it is doing.
Alternatively we may stop considering as invalid urls who do not end with a slash, as these appear to be perfectly valid for all other operations.

So, I went through both checks with the debugger. `isBaseUrlInvalid` does return true in both cases when you use a url without a final slash. However in the login process, that does not stop the login attempt. Login is successful and thus you proceed to the home page. In the source adding page, if the base url is considered invalid then it immediately kicks you out without attempting to perform any operation. It is supposedly a bit strange, because this is the only place in the application where we check the validity of the url after we already did log in. I'm not really sure the check itself is necessary and may be removed altogether from the sources. I think the check is meant to ensure we actually did log in and are not connected as a guest, but that is not really what it is doing. Alternatively we may stop considering as invalid urls who do not end with a slash, as these appear to be perfectly valid for all other operations.
AmineB commented 2025-03-09 13:43:39 +00:00
Owner
Copy Link
Copy Source

Thanks for the details. I fixed the issue.

Thanks for the details. I fixed the issue.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Difficulty - Beginner friendly
Difficulty = Easy
Difficulty = Hard
Difficulty = Medium
Priority = CRITICAL
Priority = High
Priority = Low
Priority = Normal
Priority = Some day
Status - Fixed somewhere else
Status - No details provided
Status = Can't fix
Status = Duplicate
Status = Help wanted
Status = Invalid
Status = Need more details
Status = Taken
Status = Wontfix
Type - Selfoss works like this
Type = Bug
Type = Chore
Type = Enhancement
Type = Feature
Type = Question
Type = SELFOSS API ISSUE
Type = Tools
Type = UX/Design
Up For Grabs
No labels Status = Need more details
Milestone
No items
No Milestone
Assignees
Clear assignees
AmineB (Amine Bouabdallaoui)
No Assignees
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Louvorg/ReaderForSelfoss-multiplatform#184