ui: avoid crashing when restart command is unavailable

[JR-Morton]

Summary

Fix the GTK Hang dialog so it does not offer "Relaunch this application"
when the report has no ProcCmdline.

The regular crash dialog already has this guard, but the Hang-specific
path did not. If the checkbox was selected, apport later called
UserInterface.restart(), which asserts that ProcCmdline exists.

Notes

Report.add_proc_info() normally sets ProcCmdline, but reports can still
lack it if the field is removed by hooks or if the report was generated
outside the normal collection path. In that case the UI should not offer a
restart.

Changes

• Only show the GTK Hang relaunch checkbox when ProcCmdline is present.
• Also require ProcCmdline before setting response.restart.
• Add a GTK system test covering a Hang report without ProcCmdline.
• Keep UserInterface.restart()'s assertion unchanged.

Test

• python3 -m unittest tests.unit.test_ui
• python3 -m unittest tests.system.test_ui_gtk
• git diff --check

[codecov]

Codecov Report

❌ Patch coverage is 97.36842% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 84.58%. Comparing base (b24ed44) to head (5f699c8).

Files with missing lines	Patch %	Lines
kde/apport-kde	50.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #624      +/-   ##
==========================================
+ Coverage   84.54%   84.58%   +0.03%     
==========================================
  Files         104      104              
  Lines       21028    21061      +33     
  Branches     3219     3221       +2     
==========================================
+ Hits        17779    17815      +36     
+ Misses       2809     2805       -4     
- Partials      440      441       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[bdrung]

Instead of making restart a no-op, could we change the UI instead to not allow restarting in case ProcCmdline is missing?

[schopin-pro]

Subsidiary question: how did we get an empty field in the first place? Is it a normal occurrence or an actual bug further up the data stream?

…

On Tue, 5 May 2026, 21:15 Benjamin Drung, ***@***.***> wrote: ***@***.**** commented on this pull request. Instead of making restart a no-op, could we change the UI instead to not allow restarting in case ProcCmdline is missing? — Reply to this email directly, view it on GitHub <#624 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUXZP7PFDMRTMLZ6XEV2VW34ZI4T7AVCNFSM6AAAAACYQPNINCVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHM2DEMZQHE2DEOJUHE> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[bdrung]

We need the same kind of change to kde/apport-kde.

[bdrung]

Could we extent the UI object with an can_restart method that does this check?

[bdrung]

I wasn't aware that we have offer_restart. This can make having can_restart and offer_restart confusing. The UI should only present the restart when we want to offer it plus apport can do that.