Fix test core dump, add producer metadata, real-time dashboard plugin

[Copilot]

CI crashes with SIGABRT because SKIP() and FAIL() require exceptions but the project builds with -fno-exceptions. Also addresses several infrastructure requests: producer metadata, function-oriented plugins, real-time dashboard, and file rename.

Test recovery (-fno-exceptions fix)

• Replace SKIP() with if constexpr (!enabled) { return; } — no exception needed
• Replace FAIL() in assert handler with FAIL_CHECK() (non-throwing) + longjmp recovery
• Add PLEXDB_TEST_SCOPE() macro for opt-in graceful exit on assertion failure
• Catch2 listener resets recovery state between test cases
• Enable -DPLEXDB_LOG_ENABLED=ON -DPLEXDB_DEBUG=ON in test.yml

// Before: terminates process
FAIL("Assert failed ...");

// After: records failure, longjmps back to PLEXDB_TEST_SCOPE if active
FAIL_CHECK("Assert failed ...");

Producer metadata

• New PLEXDB_LOG_PRODUCER_META = 5 event type in log_abi.h
• PlexdbLogProducerMeta { producer_id, key, value } for key-value metadata on producers
• fire_producer_meta() added to C++ module, handled in all consumers/plugins

Plugin refactor (functions over OOP)

• log_stat_plugin.cpp, log_file_plugin.cpp: classes with methods → plain structs with free functions
• log_consumer.test.cpp → log_consumer_helper.test.cpp (both plexdb and objstore), also refactored

Real-time dashboard plugin

• New objstore/plugins/log_dashboard/log_dashboard_plugin.cpp — ANSI terminal dashboard via LD_PRELOAD
• Renders live producer/stat/metadata updates, throttled by PLEXDB_DASHBOARD_INTERVAL_MS
• plot_stats.py updated with --live mode and D tag for producer metadata
• Stat plugin format extended with D <producer_id>\t<key>\t<value> lines

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Improve Testing Visibility</issue_title>
<issue_description>- Allow tests to fail an assert but move to the next test case or section

• Ensure parsing errors are correctly visible as UNSCOPED_INFO
• Add a basic zero-overhead logging system with log levels and consumers using the existing plexdb plugin infrastructure. The logs should be structured where possible and also allow realtime monitoring of statstics (without the overhead of string formatting) for realtime metrics.
• Route all logs to be visible after test failure.</issue_description>

<agent_instructions>Make the minimal possible changes, focusing on keeping the current style and quality. Ensure things are highly visibile and obvious. Update the agent instructions to ensure agents can debug the causes of test failures, ideally by enabling the agent to configure plugins for filtering. this may require introducing configuration files, but where possible, use the plugin infrastructure.</agent_instructions>

Comments on the Issue (you are @copilot in this section)

• Fixes Improve Testing Visibility #32

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.