"""
FHL Bible MCP Server - Prompts 診斷測試套件
完整的 Prompts 診斷測試，用於識別所有問題並生成診斷報告
"""
import pytest
import sys
import os
from typing import Dict, List, Tuple, Any
from dataclasses import dataclass
import json
# 添加專案根目錄到 Python path
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
from src.fhl_bible_mcp.prompts.manager import PromptManager
from src.fhl_bible_mcp.prompts.base import PromptTemplate
# ==================== 測試配置 ====================
# 長度標準（嚴格）
LENGTH_STANDARDS = {
'basic': 500, # 基礎類
'reading': 700, # 讀經類
'study': 800, # 研經類
'special': 900, # 特殊類
'advanced': 1000 # 進階類
}
# 所有 Prompts 的完整清單（19 個）
ALL_PROMPTS = [
# Basic (4)
('basic', 'basic_help_guide', 'BasicHelpGuidePrompt'),
('basic', 'basic_uri_demo', 'BasicURIDemoPrompt'),
('basic', 'basic_quick_lookup', 'BasicQuickLookupPrompt'),
('basic', 'basic_tool_reference', 'BasicToolReferencePrompt'),
# Reading (3)
('reading', 'reading_daily', 'ReadingDailyPrompt'),
('reading', 'reading_chapter', 'ReadingChapterPrompt'),
('reading', 'reading_passage', 'ReadingPassagePrompt'),
# Study (4)
('study', 'study_verse_deep', 'StudyVerseDeepPrompt'),
('study', 'study_topic_deep', 'StudyTopicDeepPrompt'),
('study', 'study_translation_compare', 'StudyTranslationComparePrompt'),
('study', 'study_word_original', 'StudyWordOriginalPrompt'),
# Special (5)
('special', 'special_sermon_prep', 'SpecialSermonPrepPrompt'),
('special', 'special_devotional', 'SpecialDevotionalPrompt'),
('special', 'special_memory_verse', 'SpecialMemoryVersePrompt'),
('special', 'special_topical_chain', 'SpecialTopicalChainPrompt'),
('special', 'special_bible_trivia', 'SpecialBibleTriviaPrompt'),
# Advanced (3)
('advanced', 'advanced_cross_reference', 'AdvancedCrossReferencePrompt'),
('advanced', 'advanced_parallel_gospels', 'AdvancedParallelGospelsPrompt'),
('advanced', 'advanced_character_study', 'AdvancedCharacterStudyPrompt'),
]
@dataclass
class PromptDiagnosticResult:
"""Prompt 診斷結果"""
category: str
name: str
class_name: str
import_success: bool
import_error: str = ""
instantiate_success: bool = False
instantiate_error: str = ""
render_success: bool = False
render_error: str = ""
rendered_length: int = 0
length_standard: int = 0
length_over: int = 0
length_status: str = "UNKNOWN" # PASS / FAIL / ERROR
has_clear_steps: bool = False
step_count: int = 0
has_action_verbs: bool = False
structure_score: int = 0 # 0-100
structure_notes: str = ""
class PromptsDiagnostics:
"""Prompts 診斷測試類"""
def __init__(self):
self.results: List[PromptDiagnosticResult] = []
self.manager = None
# ==================== 測試 1: 導入測試 ====================
def test_import_all_prompts(self) -> Tuple[int, int, List[PromptDiagnosticResult]]:
"""
測試 1: 所有 prompts 能否正確導入
Returns:
(成功數, 失敗數, 結果列表)
"""
print("\n" + "="*80)
print("測試 1: Prompts 導入測試")
print("="*80)
success_count = 0
fail_count = 0
results = []
for category, name, class_name in ALL_PROMPTS:
result = PromptDiagnosticResult(
category=category,
name=name,
class_name=class_name,
import_success=False,
length_standard=LENGTH_STANDARDS[category]
)
try:
# 嘗試導入
module_path = f"src.fhl_bible_mcp.prompts.{category}"
module = __import__(module_path, fromlist=[class_name])
prompt_class = getattr(module, class_name)
result.import_success = True
success_count += 1
print(f"✅ {name}: 導入成功")
except Exception as e:
result.import_success = False
result.import_error = str(e)
fail_count += 1
print(f"❌ {name}: 導入失敗 - {e}")
results.append(result)
print(f"\n📊 導入測試結果: {success_count}/{len(ALL_PROMPTS)} 成功")
self.results = results
return success_count, fail_count, results
# ==================== 測試 2: 實例化測試 ====================
def test_instantiate_all_prompts(self) -> Tuple[int, int]:
"""
測試 2: 所有 prompts 能否實例化
Returns:
(成功數, 失敗數)
"""
print("\n" + "="*80)
print("測試 2: Prompts 實例化測試")
print("="*80)
success_count = 0
fail_count = 0
for result in self.results:
if not result.import_success:
print(f"⏭️ {result.name}: 跳過（導入失敗）")
continue
try:
# 嘗試實例化
module_path = f"src.fhl_bible_mcp.prompts.{result.category}"
module = __import__(module_path, fromlist=[result.class_name])
prompt_class = getattr(module, result.class_name)
# 無參數實例化
instance = prompt_class()
result.instantiate_success = True
success_count += 1
print(f"✅ {result.name}: 實例化成功")
except Exception as e:
result.instantiate_success = False
result.instantiate_error = str(e)
fail_count += 1
print(f"❌ {result.name}: 實例化失敗 - {e}")
print(f"\n📊 實例化測試結果: {success_count}/{len([r for r in self.results if r.import_success])} 成功")
return success_count, fail_count
# ==================== 測試 3: 渲染測試 ====================
def test_render_all_prompts(self) -> Tuple[int, int]:
"""
測試 3: 所有 prompts 能否正常渲染
Returns:
(成功數, 失敗數)
"""
print("\n" + "="*80)
print("測試 3: Prompts 渲染測試")
print("="*80)
success_count = 0
fail_count = 0
for result in self.results:
if not result.import_success or not result.instantiate_success:
print(f"⏭️ {result.name}: 跳過（導入或實例化失敗）")
continue
try:
# 嘗試渲染
module_path = f"src.fhl_bible_mcp.prompts.{result.category}"
module = __import__(module_path, fromlist=[result.class_name])
prompt_class = getattr(module, result.class_name)
instance = prompt_class()
# 使用默認參數渲染
rendered = instance.render()
if rendered and isinstance(rendered, str):
result.render_success = True
result.rendered_length = len(rendered)
success_count += 1
print(f"✅ {result.name}: 渲染成功 ({result.rendered_length} 字)")
else:
result.render_success = False
result.render_error = "渲染返回空或非字符串"
fail_count += 1
print(f"❌ {result.name}: 渲染失敗 - 返回空或非字符串")
except Exception as e:
result.render_success = False
result.render_error = str(e)
fail_count += 1
print(f"❌ {result.name}: 渲染失敗 - {e}")
print(f"\n📊 渲染測試結果: {success_count}/{len([r for r in self.results if r.import_success and r.instantiate_success])} 成功")
return success_count, fail_count
# ==================== 測試 4: 長度分析 ====================
def test_prompts_length_analysis(self) -> Dict[str, Any]:
"""
測試 4: 長度分析
Returns:
統計數據字典
"""
print("\n" + "="*80)
print("測試 4: Prompts 長度分析")
print("="*80)
pass_count = 0
fail_count = 0
total_length = 0
max_length = 0
max_length_prompt = ""
# 按類別統計
category_stats = {
'basic': {'count': 0, 'total': 0, 'pass': 0, 'fail': 0},
'reading': {'count': 0, 'total': 0, 'pass': 0, 'fail': 0},
'study': {'count': 0, 'total': 0, 'pass': 0, 'fail': 0},
'special': {'count': 0, 'total': 0, 'pass': 0, 'fail': 0},
'advanced': {'count': 0, 'total': 0, 'pass': 0, 'fail': 0},
}
print(f"\n{'Prompt':<35} {'長度':<8} {'標準':<8} {'超標':<8} {'狀態':<8}")
print("-" * 75)
for result in self.results:
if not result.render_success:
result.length_status = "ERROR"
print(f"{result.name:<35} {'N/A':<8} {result.length_standard:<8} {'N/A':<8} ❌ ERROR")
continue
# 計算超標量
result.length_over = result.rendered_length - result.length_standard
# 判斷狀態
if result.rendered_length <= result.length_standard:
result.length_status = "PASS"
pass_count += 1
category_stats[result.category]['pass'] += 1
status_icon = "✅ PASS"
else:
result.length_status = "FAIL"
fail_count += 1
category_stats[result.category]['fail'] += 1
status_icon = "❌ FAIL"
# 統計
total_length += result.rendered_length
category_stats[result.category]['count'] += 1
category_stats[result.category]['total'] += result.rendered_length
if result.rendered_length > max_length:
max_length = result.rendered_length
max_length_prompt = result.name
# 打印結果
print(f"{result.name:<35} {result.rendered_length:<8} {result.length_standard:<8} {result.length_over if result.length_over > 0 else '-':<8} {status_icon}")
# 計算平均值
render_success_count = len([r for r in self.results if r.render_success])
avg_length = total_length / render_success_count if render_success_count > 0 else 0
# 打印統計
print("\n" + "="*80)
print("📊 長度統計")
print("="*80)
print(f"總計: {render_success_count} 個 prompts")
print(f"通過: {pass_count} 個 ({pass_count/render_success_count*100:.1f}%)")
print(f"未通過: {fail_count} 個 ({fail_count/render_success_count*100:.1f}%)")
print(f"平均長度: {avg_length:.0f} 字")
print(f"最長 Prompt: {max_length_prompt} ({max_length} 字)")
print("\n按類別統計:")
print("-" * 75)
print(f"{'類別':<12} {'數量':<8} {'平均長度':<12} {'通過':<8} {'未通過':<8}")
print("-" * 75)
for category, stats in category_stats.items():
if stats['count'] > 0:
avg = stats['total'] / stats['count']
print(f"{category:<12} {stats['count']:<8} {avg:<12.0f} {stats['pass']:<8} {stats['fail']:<8}")
return {
'total': render_success_count,
'pass': pass_count,
'fail': fail_count,
'avg_length': avg_length,
'max_length': max_length,
'max_length_prompt': max_length_prompt,
'category_stats': category_stats
}
# ==================== 測試 5: 結構檢查 ====================
def test_prompts_structure_check(self) -> Dict[str, Any]:
"""
測試 5: 結構檢查
檢查 prompt 是否包含清晰的步驟、動詞開頭的指令等
Returns:
統計數據字典
"""
print("\n" + "="*80)
print("測試 5: Prompts 結構檢查")
print("="*80)
action_verbs = [
'查詢', '搜尋', '分析', '比較', '列出', '提取', '總結', '歸納',
'檢查', '驗證', '計算', '統計', '排序', '篩選', '組織', '建立',
'創建', '生成', '輸出', '顯示', '展示', '使用', '執行', '調用'
]
good_structure = 0
poor_structure = 0
print(f"\n{'Prompt':<35} {'步驟':<8} {'動詞':<8} {'評分':<8} {'狀態':<10}")
print("-" * 75)
for result in self.results:
if not result.render_success:
print(f"{result.name:<35} {'N/A':<8} {'N/A':<8} {'N/A':<8} ⏭️ ERROR")
continue
# 獲取渲染內容
try:
module_path = f"src.fhl_bible_mcp.prompts.{result.category}"
module = __import__(module_path, fromlist=[result.class_name])
prompt_class = getattr(module, result.class_name)
instance = prompt_class()
rendered = instance.render()
# 檢查步驟標記
step_markers = ['## 步驟', '步驟 1', '步驟 2', '步驟 3', 'Step 1', 'Step 2']
result.has_clear_steps = any(marker in rendered for marker in step_markers)
# 計算步驟數
result.step_count = sum(rendered.count(f'步驟 {i}') for i in range(1, 20))
if result.step_count == 0:
result.step_count = sum(rendered.count(f'Step {i}') for i in range(1, 20))
# 檢查動詞使用
verb_count = sum(rendered.count(verb) for verb in action_verbs)
result.has_action_verbs = verb_count >= 3
# 計算結構評分 (0-100)
score = 0
notes = []
# 有清晰步驟 +40
if result.has_clear_steps:
score += 40
notes.append("有步驟標記")
else:
notes.append("缺步驟標記")
# 步驟數量 3-7 個 +30
if 3 <= result.step_count <= 7:
score += 30
notes.append(f"步驟數適中({result.step_count})")
elif result.step_count > 0:
score += 15
notes.append(f"步驟數欠佳({result.step_count})")
else:
notes.append("無明確步驟")
# 有動詞 +30
if result.has_action_verbs:
score += 30
notes.append("有動作動詞")
else:
notes.append("缺動作動詞")
result.structure_score = score
result.structure_notes = "; ".join(notes)
# 判斷結構品質
if score >= 70:
status = "✅ 良好"
good_structure += 1
else:
status = "⚠️ 待改善"
poor_structure += 1
# 打印結果
step_str = str(result.step_count) if result.step_count > 0 else '-'
verb_str = '✓' if result.has_action_verbs else '✗'
print(f"{result.name:<35} {step_str:<8} {verb_str:<8} {score:<8} {status}")
except Exception as e:
result.structure_notes = f"檢查失敗: {e}"
print(f"{result.name:<35} {'N/A':<8} {'N/A':<8} {'N/A':<8} ❌ ERROR")
# 統計
render_success_count = len([r for r in self.results if r.render_success])
print("\n" + "="*80)
print("📊 結構統計")
print("="*80)
print(f"良好結構: {good_structure}/{render_success_count} ({good_structure/render_success_count*100:.1f}%)")
print(f"待改善: {poor_structure}/{render_success_count} ({poor_structure/render_success_count*100:.1f}%)")
return {
'good_structure': good_structure,
'poor_structure': poor_structure,
'total': render_success_count
}
# ==================== 測試 6: PromptManager 整合 ====================
def test_prompt_manager_integration(self) -> Tuple[bool, str]:
"""
測試 6: PromptManager 整合
Returns:
(成功與否, 訊息)
"""
print("\n" + "="*80)
print("測試 6: PromptManager 整合測試")
print("="*80)
try:
# 創建 PromptManager
manager = PromptManager()
self.manager = manager
# 檢查註冊數量
registered_count = len(manager.get_prompt_names())
expected_count = 19
print(f"\n註冊 Prompts 數量: {registered_count}")
print(f"預期數量: {expected_count}")
if registered_count == expected_count:
print(f"✅ PromptManager 整合成功")
# 測試 list_prompts
prompts_list = manager.list_prompts()
print(f"\nlist_prompts() 返回: {len(prompts_list)} 個 prompts")
# 測試 get_prompt
sample_prompt = manager.get_prompt('basic_help_guide')
if sample_prompt:
print(f"✅ get_prompt() 成功獲取 prompt")
else:
print(f"⚠️ get_prompt() 無法獲取 prompt")
return True, "PromptManager 整合測試通過"
else:
msg = f"❌ 註冊數量不符: 預期 {expected_count}, 實際 {registered_count}"
print(msg)
return False, msg
except Exception as e:
msg = f"❌ PromptManager 整合失敗: {e}"
print(msg)
return False, msg
# ==================== 生成報告 ====================
def generate_report(self, output_path: str = None) -> str:
"""
生成診斷報告
Args:
output_path: 報告輸出路徑，默認為 docs/PROMPTS_DIAGNOSTIC_REPORT.md
Returns:
報告內容
"""
if output_path is None:
output_path = os.path.join(
os.path.dirname(__file__),
'..',
'docs',
'PROMPTS_DIAGNOSTIC_REPORT.md'
)
# 統計數據
total = len(self.results)
import_success = len([r for r in self.results if r.import_success])
instantiate_success = len([r for r in self.results if r.instantiate_success])
render_success = len([r for r in self.results if r.render_success])
length_pass = len([r for r in self.results if r.length_status == "PASS"])
length_fail = len([r for r in self.results if r.length_status == "FAIL"])
# 分類問題
p0_issues = [r for r in self.results if not r.import_success or not r.instantiate_success or not r.render_success]
p1_issues = [r for r in self.results if r.render_success and r.length_over > r.length_standard * 0.5]
p2_issues = [r for r in self.results if r.render_success and 0 < r.length_over <= r.length_standard * 0.5]
p3_issues = [r for r in self.results if r.render_success and r.structure_score < 70]
# 生成報告
report = f"""# Prompts 診斷報告 🔍
**FHL Bible MCP Server - Prompts 優化診斷**
---
## 執行資訊
- **執行時間**: {self._get_current_time()}
- **測試版本**: 1.0
- **測試 Prompts 數量**: {total}
- **測試項目**: 6 項（導入、實例化、渲染、長度、結構、整合）
---
## 📊 測試結果總覽
### 基本測試
| 測試項目 | 成功 | 失敗 | 成功率 |
|---------|------|------|--------|
| **導入測試** | {import_success}/{total} | {total - import_success}/{total} | {import_success/total*100:.1f}% |
| **實例化測試** | {instantiate_success}/{total} | {total - instantiate_success}/{total} | {instantiate_success/total*100:.1f}% |
| **渲染測試** | {render_success}/{total} | {total - render_success}/{total} | {render_success/total*100:.1f}% |
### 長度分析
| 測試項目 | 通過 | 未通過 | 通過率 |
|---------|------|--------|--------|
| **長度標準** | {length_pass}/{render_success} | {length_fail}/{render_success} | {length_pass/render_success*100 if render_success > 0 else 0:.1f}% |
**標準**: 基礎 < 500 字，讀經 < 700 字，研經 < 800 字，特殊 < 900 字，進階 < 1000 字
---
## 📋 詳細分析
### 長度分析表
| Prompt | 類別 | 當前長度 | 標準 | 超標 | 狀態 |
|--------|------|----------|------|------|------|
"""
# 添加每個 prompt 的詳細資料
for result in sorted(self.results, key=lambda r: r.length_over if r.render_success else 0, reverse=True):
if result.render_success:
status_icon = "✅" if result.length_status == "PASS" else "❌"
over_str = f"+{result.length_over}" if result.length_over > 0 else "-"
report += f"| `{result.name}` | {result.category} | {result.rendered_length} | {result.length_standard} | {over_str} | {status_icon} {result.length_status} |\n"
else:
report += f"| `{result.name}` | {result.category} | N/A | {result.length_standard} | N/A | ❌ ERROR |\n"
report += f"""
### 結構分析表
| Prompt | 步驟數 | 結構評分 | 狀態 | 備註 |
|--------|--------|----------|------|------|
"""
for result in sorted(self.results, key=lambda r: r.structure_score if r.render_success else 0):
if result.render_success:
status_icon = "✅" if result.structure_score >= 70 else "⚠️"
step_str = str(result.step_count) if result.step_count > 0 else "-"
report += f"| `{result.name}` | {step_str} | {result.structure_score}/100 | {status_icon} | {result.structure_notes} |\n"
else:
report += f"| `{result.name}` | - | - | ❌ | 渲染失敗 |\n"
report += f"""
---
## 🚨 問題分類
### P0: 載入/渲染失敗（必須立即修復）
**數量**: {len(p0_issues)} 個
"""
if p0_issues:
for result in p0_issues:
report += f"""#### `{result.name}`
- **類別**: {result.category}
- **問題**:
"""
if not result.import_success:
report += f" - ❌ 導入失敗: `{result.import_error}`\n"
if not result.instantiate_success:
report += f" - ❌ 實例化失敗: `{result.instantiate_error}`\n"
if not result.render_success:
report += f" - ❌ 渲染失敗: `{result.render_error}`\n"
report += "\n"
else:
report += "✅ 無 P0 問題\n\n"
report += f"""### P1: 嚴重超長（高優先級）
**數量**: {len(p1_issues)} 個
**標準**: 超過標準 50% 以上
"""
if p1_issues:
for result in sorted(p1_issues, key=lambda r: r.length_over, reverse=True):
over_pct = (result.length_over / result.length_standard) * 100
report += f"- [ ] **`{result.name}`**: {result.rendered_length} 字（標準 {result.length_standard}，超標 {result.length_over} 字，+{over_pct:.0f}%）\n"
report += "\n"
else:
report += "✅ 無 P1 問題\n\n"
report += f"""### P2: 中度超長（中優先級）
**數量**: {len(p2_issues)} 個
**標準**: 超過標準 0-50%
"""
if p2_issues:
for result in sorted(p2_issues, key=lambda r: r.length_over, reverse=True):
over_pct = (result.length_over / result.length_standard) * 100
report += f"- [ ] **`{result.name}`**: {result.rendered_length} 字（標準 {result.length_standard}，超標 {result.length_over} 字，+{over_pct:.0f}%）\n"
report += "\n"
else:
report += "✅ 無 P2 問題\n\n"
report += f"""### P3: 結構待優化（低優先級）
**數量**: {len(p3_issues)} 個
**標準**: 結構評分 < 70
"""
if p3_issues:
for result in sorted(p3_issues, key=lambda r: r.structure_score):
report += f"- [ ] **`{result.name}`**: 評分 {result.structure_score}/100（{result.structure_notes}）\n"
report += "\n"
else:
report += "✅ 無 P3 問題\n\n"
report += f"""---
## 📈 統計數據
### 按類別統計
"""
# 按類別統計
category_summary = {}
for result in self.results:
if result.category not in category_summary:
category_summary[result.category] = {
'total': 0,
'render_success': 0,
'length_pass': 0,
'length_fail': 0,
'total_length': 0,
'avg_length': 0,
'standard': result.length_standard
}
category_summary[result.category]['total'] += 1
if result.render_success:
category_summary[result.category]['render_success'] += 1
category_summary[result.category]['total_length'] += result.rendered_length
if result.length_status == "PASS":
category_summary[result.category]['length_pass'] += 1
elif result.length_status == "FAIL":
category_summary[result.category]['length_fail'] += 1
# 計算平均值
for cat, stats in category_summary.items():
if stats['render_success'] > 0:
stats['avg_length'] = stats['total_length'] / stats['render_success']
report += "| 類別 | 數量 | 標準 | 平均長度 | 通過 | 未通過 | 通過率 |\n"
report += "|------|------|------|----------|------|--------|--------|\n"
for cat in ['basic', 'reading', 'study', 'special', 'advanced']:
if cat in category_summary:
stats = category_summary[cat]
pass_rate = (stats['length_pass'] / stats['render_success'] * 100) if stats['render_success'] > 0 else 0
report += f"| {cat} | {stats['total']} | {stats['standard']} | {stats['avg_length']:.0f} | {stats['length_pass']} | {stats['length_fail']} | {pass_rate:.1f}% |\n"
report += f"""
### 總體統計
- **總 Prompts**: {total}
- **渲染成功**: {render_success}
- **長度通過**: {length_pass}
- **長度未通過**: {length_fail}
- **平均長度**: {sum(r.rendered_length for r in self.results if r.render_success) / render_success if render_success > 0 else 0:.0f} 字
- **最長 Prompt**: {max((r for r in self.results if r.render_success), key=lambda r: r.rendered_length).name if render_success > 0 else 'N/A'} ({max((r.rendered_length for r in self.results if r.render_success), default=0)} 字)
---
## 🎯 重構建議
### 優先級排序
1. **立即修復 P0** ({len(p0_issues)} 個)
- 修復導入/實例化/渲染錯誤
- 確保所有 prompts 都能正常載入
2. **高優先級 P1** ({len(p1_issues)} 個)
- 嚴重超長的 prompts
- 需要大幅縮減內容（50%+ 超標）
3. **中優先級 P2** ({len(p2_issues)} 個)
- 中度超長的 prompts
- 需要適度縮減內容（0-50% 超標）
4. **低優先級 P3** ({len(p3_issues)} 個)
- 結構待優化的 prompts
- 改善步驟清晰度和動詞使用
### 重構策略
**漸進優化（Strategy B）**:
1. Phase 1: 修復 P0 問題（Day 1-2）
2. Phase 2: 重構 P1 問題（Day 3-7）
3. Phase 3: 優化 P2 問題（Day 8-10）
4. Phase 4: 改善 P3 問題（Day 11-12）
### 目標
- ✅ **100% 載入成功率**
- ✅ **100% 長度合規**
- ✅ **90%+ 結構良好**
- ✅ **平均長度 < 700 字**
---
## 📝 下一步行動
1. **審閱報告**: 確認問題分類和優先級
2. **準備重構**: 創建重構模板和檢查清單
3. **開始修復**: 從 P0 問題開始，逐步解決
4. **測試驗證**: 每次重構後運行診斷測試
5. **更新文檔**: 將詳細說明移至使用指南
---
**報告生成時間**: {self._get_current_time()}
**報告版本**: 1.0
**狀態**: 診斷完成，待重構
---
Made with ❤️ for better Prompts 🚀
"""
# 寫入檔案
os.makedirs(os.path.dirname(output_path), exist_ok=True)
with open(output_path, 'w', encoding='utf-8') as f:
f.write(report)
print(f"\n✅ 診斷報告已生成: {output_path}")
return report
def _get_current_time(self) -> str:
"""獲取當前時間字串"""
from datetime import datetime
return datetime.now().strftime("%Y-%m-%d %H:%M:%S")
# ==================== Pytest 測試函數 ====================
@pytest.fixture(scope="module")
def diagnostics():
"""創建診斷測試實例"""
return PromptsDiagnostics()
def test_1_import_all_prompts(diagnostics):
"""測試 1: 導入測試"""
success, fail, results = diagnostics.test_import_all_prompts()
assert success > 0, "至少應該有一個 prompt 導入成功"
def test_2_instantiate_all_prompts(diagnostics):
"""測試 2: 實例化測試"""
success, fail = diagnostics.test_instantiate_all_prompts()
assert success > 0, "至少應該有一個 prompt 實例化成功"
def test_3_render_all_prompts(diagnostics):
"""測試 3: 渲染測試"""
success, fail = diagnostics.test_render_all_prompts()
assert success > 0, "至少應該有一個 prompt 渲染成功"
def test_4_prompts_length_analysis(diagnostics):
"""測試 4: 長度分析"""
stats = diagnostics.test_prompts_length_analysis()
assert stats['total'] > 0, "應該有可分析的 prompts"
def test_5_prompts_structure_check(diagnostics):
"""測試 5: 結構檢查"""
stats = diagnostics.test_prompts_structure_check()
assert stats['total'] > 0, "應該有可檢查的 prompts"
def test_6_prompt_manager_integration(diagnostics):
"""測試 6: PromptManager 整合"""
success, msg = diagnostics.test_prompt_manager_integration()
assert success, f"PromptManager 整合失敗: {msg}"
def test_7_generate_report(diagnostics):
"""測試 7: 生成診斷報告"""
report = diagnostics.generate_report()
assert report and len(report) > 0, "報告應該有內容"
assert "# Prompts 診斷報告" in report, "報告應該包含標題"
# ==================== 主程序 ====================
if __name__ == "__main__":
print("🔍 FHL Bible MCP Server - Prompts 診斷測試")
print("=" * 80)
diagnostics = PromptsDiagnostics()
# 執行所有測試
diagnostics.test_import_all_prompts()
diagnostics.test_instantiate_all_prompts()
diagnostics.test_render_all_prompts()
diagnostics.test_prompts_length_analysis()
diagnostics.test_prompts_structure_check()
diagnostics.test_prompt_manager_integration()
# 生成報告
print("\n" + "="*80)
print("生成診斷報告...")
print("="*80)
diagnostics.generate_report()
print("\n✅ 診斷測試完成！")
print("\n請查看 docs/PROMPTS_DIAGNOSTIC_REPORT.md 獲取詳細報告")
