Add PersistentShell fallback to avoid getting stuck at PersistentShell

[GT-610]

Resolve #1142.

Summary by CodeRabbit

• Bug Fixes
  • More reliable server status checks: uses a persistent connection by default and automatically falls back to a direct SSH path on timeouts.
  • Improved timeout handling and recovery to prevent stale connections and ensure retries complete.
  • Better Windows handling for status retrieval and clearer warnings when no SSH client is available.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: a93d99ae-a286-4f63-9e59-4638dea406d9

📥 Commits

Reviewing files that changed from the base of the PR and between 2ac9ba4 and 4678daa.

📒 Files selected for processing (1)

• lib/data/provider/server/single.dart

🚧 Files skipped from review as they are similar to previous changes (1)

• lib/data/provider/server/single.dart

---

📝 Walkthrough

Walkthrough

This change refactors server status command execution in ServerNotifier to use persistent shell with fallback. A new private flag controls whether persistent-shell execution is attempted; it defaults to true and resets when the SSH client instance changes. Two helper methods (_runStatusCommand and _runStatusCommandWithExec) implement the strategy: persistent shell is attempted first but on timeout the flag is disabled, the shell is disposed, and the command is retried via exec. _getData now delegates status execution to _runStatusCommand. Windows status commands are forced through the exec path.

Possibly related PRs

• lollipopkit/flutter_server_box#1124: Modifies server status execution to create/dispose persistent shells, handle timeouts, and fall back to exec similarly.

Suggested reviewers

• lollipopkit

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title directly addresses the main change: implementing a PersistentShell fallback mechanism to prevent the connection flow from getting stuck on PersistentShell timeouts.
Linked Issues check	✅ Passed	The code changes implement a fallback mechanism from PersistentShell to direct SSH exec when timeouts occur, directly addressing the issue #1142 requirement to restore server-page connectivity without requiring server_box_monitor.
Out of Scope Changes check	✅ Passed	All changes are scoped to implementing the PersistentShell fallback mechanism in ServerNotifier; no extraneous modifications detected.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix-persistent-shell-fallback

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

lib/data/provider/server/single.dart (1)

529-532: 💤 Low value

Misleading log message when client is null.

The message "No status result from ${spi.name}" implies a failed status fetch, but at this point no attempt was made—the client is simply unavailable.

Suggested improvement

    if (client == null) {
-     Loggers.app.warning('No status result from ${spi.name}');
+     Loggers.app.warning('SSH client not available for ${spi.name}');
      return '';
    }

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@lib/data/provider/server/single.dart` around lines 529 - 532, The warning
message is misleading because no status fetch was attempted when client == null;
update the Loggers.app.warning call (the null-check handling for variable client
in this function) to clearly state that the client is unavailable (e.g., "No
client available for ${spi.name}" or "Client for ${spi.name} is null, skipping
status fetch") while preserving the existing return '' behavior; change only the
log text invoked via Loggers.app.warning and keep the client null-check and
return as-is.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@lib/data/provider/server/single.dart`:
- Around line 529-532: The warning message is misleading because no status fetch
was attempted when client == null; update the Loggers.app.warning call (the
null-check handling for variable client in this function) to clearly state that
the client is unavailable (e.g., "No client available for ${spi.name}" or
"Client for ${spi.name} is null, skipping status fetch") while preserving the
existing return '' behavior; change only the log text invoked via
Loggers.app.warning and keep the client null-check and return as-is.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 06432391-0208-4c74-b5a4-8ac5c5395afa

📥 Commits

Reviewing files that changed from the base of the PR and between 6d4873d and 2ac9ba4.

📒 Files selected for processing (1)

• lib/data/provider/server/single.dart