_test_presets.py

"""Preset evaluation script for Mimir's Memory Hub.

Tests that each preset (companion, agent, character, writer, assistant, custom) produces
behaviour appropriate for its role by checking system-prompt construction, memory
tag generation, and response tone across a small set of role-specific scenarios.

Usage:
    python _test_presets.py              # quick in-process checks (no LLM needed)
    python _test_presets.py --live       # also test against a running Mimir instance
"""

import sys, json, importlib, re

# ── In-process structural tests ──────────────────────────────────────────────

def test_preset_definitions():
    """Verify every preset has required keys and sensible values."""
    sys.path.insert(0, ".")
    from playground.presets import PRESETS

    REQUIRED_KEYS = {
        "label", "description", "icon", "chemistry", "emotion_weight",
        "social_priority", "task_priority", "system_prompt_suffix",
        "mimir_overrides",
    }
    passed = 0
    for name, preset in PRESETS.items():
        missing = REQUIRED_KEYS - set(preset.keys())
        assert not missing, f"Preset '{name}' missing keys: {missing}"
        assert isinstance(preset["emotion_weight"], (int, float))
        assert 0.0 <= preset["emotion_weight"] <= 1.0, f"{name} emotion_weight out of range"
        assert isinstance(preset["system_prompt_suffix"], str) and len(preset["system_prompt_suffix"]) > 10
        passed += 1
    print(f"[PASS] All {passed} presets have valid structure")


def test_no_copilot_preset():
    """Copilot preset should have been merged into agent."""
    from playground.presets import PRESETS
    assert "copilot" not in PRESETS, "copilot preset still exists — should be merged into agent"
    print("[PASS] No copilot preset (merged into agent)")


def test_agent_has_tools():
    """Agent preset must include tool documentation."""
    from playground.presets import PRESETS
    suffix = PRESETS["agent"]["system_prompt_suffix"]
    for tool in ["read_file", "write_file", "run_code", "shell_exec", "fetch_page",
                 "http_request", "json_parse", "screenshot", "clipboard", "system_info",
                 "diff_files", "pdf_read", "csv_query", "regex_replace"]:
        assert tool in suffix, f"Agent preset missing tool doc for '{tool}'"
    assert "save_file" in suffix.lower(), "Agent preset missing save_file instructions"
    assert "workflow" in suffix.lower(), "Agent preset missing workflow section"
    print("[PASS] Agent preset has full tool documentation and workflow")


def test_companion_emotional():
    """Companion preset must prioritise emotion and social connection."""
    from playground.presets import PRESETS
    p = PRESETS["companion"]
    assert p["emotion_weight"] >= 0.7, f"Companion emotion_weight too low: {p['emotion_weight']}"
    assert p["chemistry"] is True, "Companion should enable chemistry"
    assert p["social_priority"] is True
    assert p["task_priority"] is False
    suffix = p["system_prompt_suffix"]
    assert "emotion" in suffix.lower() or "feel" in suffix.lower(), "Companion prompt lacks emotional language"
    print("[PASS] Companion preset is emotionally tuned")


def test_character_full_roleplay():
    """Character preset must maximise emotion and immersion."""
    from playground.presets import PRESETS
    p = PRESETS["character"]
    assert p["emotion_weight"] >= 0.9, f"Character emotion_weight too low: {p['emotion_weight']}"
    assert p["chemistry"] is True
    print("[PASS] Character preset is fully immersive")


def test_writer_creative():
    """Writer preset must enable creativity with project tracking."""
    from playground.presets import PRESETS
    p = PRESETS["writer"]
    assert p["emotion_weight"] >= 0.4, f"Writer emotion_weight too low: {p['emotion_weight']}"
    assert p["chemistry"] is True, "Writer should enable chemistry for creative energy"
    assert p["task_priority"] is True, "Writer should track chapters/projects"
    suffix = p["system_prompt_suffix"].lower()
    assert "creative" in suffix or "writing" in suffix, "Writer prompt lacks creative language"
    assert "chapter" in suffix or "project" in suffix, "Writer prompt should mention project tracking"
    assert "character" in suffix, "Writer prompt should mention character tracking"
    print("[PASS] Writer preset is creatively tuned with project tracking")


def test_assistant_neutral():
    """Assistant preset must be a virtual PA with tool access."""
    from playground.presets import PRESETS
    p = PRESETS["assistant"]
    assert p["emotion_weight"] <= 0.2
    assert p["task_priority"] is True
    assert p["chemistry"] is False
    suffix = p["system_prompt_suffix"].lower()
    assert "appointment" in suffix or "schedule" in suffix, "Assistant should mention scheduling"
    assert "email" in suffix, "Assistant should mention email handling"
    assert "read_file" in suffix or "pdf_read" in suffix, "Assistant should have tool docs"
    print("[PASS] Assistant preset is a virtual PA with tools")


def test_agent_task_focused():
    """Agent preset must be task-focused with low emotion."""
    from playground.presets import PRESETS
    p = PRESETS["agent"]
    assert p["emotion_weight"] <= 0.3
    assert p["task_priority"] is True
    assert p["chemistry"] is False
    print("[PASS] Agent preset is task-focused")


def test_memory_format_differentiation():
    """Companion/character use emotional memory; agent/assistant use task memory."""
    from playground.presets import PRESETS

    emotional_keywords = ["emotion", "cherish", "anchor", "social impression"]
    task_keywords = ["task", "goal", "solution", "lesson"]

    for name in ("companion", "character", "writer"):
        suffix = PRESETS[name]["system_prompt_suffix"].lower()
        found = [k for k in emotional_keywords if k in suffix]
        assert len(found) >= 2, f"{name} missing emotional memory keywords (found: {found})"

    for name in ("agent", "assistant"):
        suffix = PRESETS[name]["system_prompt_suffix"].lower()
        found = [k for k in task_keywords if k in suffix]
        assert len(found) >= 2, f"{name} missing task memory keywords (found: {found})"

    print("[PASS] Memory formats match preset roles")


def test_ui_has_no_copilot_option():
    """Check that copilot option was removed from HTML and JS."""
    with open("playground/static/index.html", "r", encoding="utf-8") as f:
        html = f.read()
    assert "copilot" not in html.lower().split("<!--")[0], "index.html still has copilot option"
    assert "writer" in html.lower(), "index.html missing writer preset option"

    with open("playground/static/js/agents.js", "r", encoding="utf-8") as f:
        js = f.read()
    assert "copilot" not in js.lower(), "agents.js still has copilot option"
    assert "writer" in js.lower(), "agents.js missing writer preset option"

    print("[PASS] UI dropdowns have no copilot, has writer")


# ── Live API tests (optional, requires running server) ───────────────────────

def test_live_presets():
    """Hit the running Mimir server to verify presets respond differently."""
    import urllib.request

    base = "http://localhost:5000"
    try:
        urllib.request.urlopen(f"{base}/api/status", timeout=3)
    except Exception:
        print("[SKIP] Live tests — server not running on :5000")
        return

    presets_resp = urllib.request.urlopen(f"{base}/api/presets")
    presets = json.loads(presets_resp.read())
    names = [p["name"] if isinstance(p, dict) else p for p in presets]
    assert "copilot" not in names, "Server still returns copilot preset"
    assert "agent" in names
    print("[PASS] /api/presets has no copilot, has agent")


# ── Runner ───────────────────────────────────────────────────────────────────

def main():
    import os
    os.chdir(os.path.dirname(os.path.abspath(__file__)))

    tests = [
        test_preset_definitions,
        test_no_copilot_preset,
        test_agent_has_tools,
        test_companion_emotional,
        test_character_full_roleplay,
        test_writer_creative,
        test_assistant_neutral,
        test_agent_task_focused,
        test_memory_format_differentiation,
        test_ui_has_no_copilot_option,
    ]
    if "--live" in sys.argv:
        tests.append(test_live_presets)

    failed = 0
    for t in tests:
        try:
            t()
        except AssertionError as e:
            print(f"[FAIL] {t.__name__}: {e}")
            failed += 1
        except Exception as e:
            print(f"[ERROR] {t.__name__}: {e}")
            failed += 1

    print(f"\n{'='*50}")
    print(f"Results: {len(tests) - failed}/{len(tests)} passed" + (" ✓" if not failed else f", {failed} failed"))
    return failed


if __name__ == "__main__":
    sys.exit(main())