Avoid progress bar overlap

[TwoOfTwelve]

Avoids overlaps of the progress bar with log outputs by delaying the log outputs while a progress bar is being shown.

Log outputs that happen during the progress bar are instead shown below it, after it is done:

Loading Submissions   100% [=========================] 2/2 (0:00:00 / 0:00:00) 
Parsing Submissions     0% [                               ] 0/2 (0:00:00 / ?) line 5:12 mismatched input '=' expecting '{'
line 5:12 mismatched input '=' expecting '{'
Parsing Submissions   100% [=========================] 2/2 (0:00:00 / 0:00:00) 
2024-07-11-12:01:45_008 [WARN] Submission - Failed to parse submission test1.cs:
failed to parse 'test1.cs'class de.jplag.csharp.grammar.CSharpParser$Namespace_or_type_nameContext cannot be cast to class de.jplag.csharp.grammar.CSharpParser$Local_variable_declarationContext (de.jplag.csharp.grammar.CSharpParser$Namespace_or_type_nameContext and de.jplag.csharp.grammar.CSharpParser$Local_variable_declarationContext are in unnamed module of loader 'app')
2024-07-11-12:01:45_012 [WARN] Submission - Failed to parse submission test2.cs:
failed to parse 'test2.cs'class de.jplag.csharp.grammar.CSharpParser$Namespace_or_type_nameContext cannot be cast to class de.jplag.csharp.grammar.CSharpParser$Local_variable_declarationContext (de.jplag.csharp.grammar.CSharpParser$Namespace_or_type_nameContext and de.jplag.csharp.grammar.CSharpParser$Local_variable_declarationContext are in unnamed module of loader 'app')

There is still some overlap with antlr, because antlr writes directly to System.err. That problem is addressed in #1868.

[tsaglam]

There are a few issues regarding sync and naming here. Also, please check the Sonar issues.

[Kr0nox]

Do we want the errors to be printed like they are in the screenshot or after a specific progress bar has run to 100%.
With the current way, a specific progress bar may be printed multiple times with different percentages.

[tsaglam]

Do we want the errors to be printed like they are in the screenshot or after a specific progress bar has run to 100%. With the current way, a specific progress bar may be printed multiple times with different percentages.

Both would be possible. Without the duplication, it looks cleaner. With the duplication, you get feedback earlier. For large datasets, this allows you to already check out the erroneous submissions while JPlag is still running.

@jplag/hiwi @jplag/maintainer any preferences?

[Kr0nox]

I think printing them after (and thus having no duplication) would be better.

1. Then all errors are accumulated together and not split between multiple instances of the progress bar
2. It is easier to follow the process of the current taks, when you dont have to switch progress bars.

But that is just my opinion. Feel free to object

[Kr0nox]

There also seems to be an issue with printing. These two failed submission are exact copies of each other, but the reason is only printed for one of them

[TwoOfTwelve]

Can you try again. Now there should be nothing cut of.

[sonarcloud]

Quality Gate passed for 'JPlag Plagiarism Detector'

Issues
0 New issues
1 Accepted issue

Measures
0 Security Hotspots
81.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

[Kr0nox]

There also seems to be an issue with printing. These two failed submission are exact copies of each other, but the reason is only printed for one of them

This seems to be fixed now

Do we want the errors to be printed like they are in the screenshot or after a specific progress bar has run to 100%.
With the current way, a specific progress bar may be printed multiple times with different percentages.

@TwoOfTwelve Did you also change it to this behavior? Or was it the default in your PR anyways or is my dataset just small enough that the errors are printed so late?

[TwoOfTwelve]

With the way it's implemented in this PR the error will always be printed after the progress bar has reach 100% or JPlag was aborted. I could try to change that behavior by interrupting the progress bar when an error occurs, but that would be a lot more work.