"""
Requirement Analyzer - 需求模糊點分析器
檢測用戶需求中的模糊點、邏輯漏洞和潛在災難場景，
決定是否需要觸發蘇格拉底面試。
"""
import re
from typing import Dict, List, Any, Optional
class RequirementAnalyzer:
"""需求分析器 - 檢測模糊點和邏輯漏洞"""
# 模糊關鍵詞（觸發面試的信號）
FUZZY_KEYWORDS = [
# 數量不明確
'一些', '幾個', '多個', '很多',
# 時間不明確
'快速', '實時', '即時', '盡快',
# 規模不明確
'大量', '海量', '高並發', '大規模',
# 邏輯不明確
'可能', '也許', '應該', '大概',
# 範圍不明確
'等等', '之類', '相關', '類似'
]
# 災難場景關鍵詞（必須追問的領域）
DISASTER_KEYWORDS = [
# 併發問題
'併發', '並發', '多用戶', '同時',
# 一致性問題
'數據', '資料', '狀態', '交易',
# 可用性問題
'支付', '訂單', '金錢', '庫存',
# 安全問題
'用戶', '認證', '權限', '密碼',
# 性能問題
'查詢', '搜索', '列表', '分頁'
]
def analyze(self, requirement: str) -> Dict[str, Any]:
"""
分析需求，返回分析結果
Args:
requirement: 用戶需求描述
Returns:
{
"needs_interview": bool,
"fuzzy_points": [...],
"disaster_scenarios": [...],
"complexity_score": 0-10
}
"""
fuzzy_points = self._detect_fuzzy_keywords(requirement)
disaster_scenarios = self._detect_disaster_scenarios(requirement)
complexity_score = self._calculate_complexity(requirement)
# 決策邏輯：有模糊點 OR 有災難場景 OR 複雜度 > 5
needs_interview = (
len(fuzzy_points) > 0 or
len(disaster_scenarios) > 0 or
complexity_score > 5
)
return {
"needs_interview": needs_interview,
"fuzzy_points": fuzzy_points,
"disaster_scenarios": disaster_scenarios,
"complexity_score": complexity_score,
"word_count": len(requirement),
"analysis_summary": self._generate_summary(
fuzzy_points,
disaster_scenarios,
complexity_score
)
}
def _detect_fuzzy_keywords(self, requirement: str) -> List[Dict[str, str]]:
"""檢測模糊關鍵詞"""
found = []
for keyword in self.FUZZY_KEYWORDS:
if keyword in requirement:
# 找到上下文
pattern = f".{{0,20}}{re.escape(keyword)}.{{0,20}}"
matches = re.finditer(pattern, requirement)
for match in matches:
found.append({
"keyword": keyword,
"context": match.group(0),
"type": "fuzzy"
})
return found
def _detect_disaster_scenarios(self, requirement: str) -> List[Dict[str, str]]:
"""檢測災難場景關鍵詞"""
found = []
for keyword in self.DISASTER_KEYWORDS:
if keyword in requirement:
# 確定災難類型
disaster_type = self._classify_disaster(keyword)
found.append({
"keyword": keyword,
"type": disaster_type,
"severity": "high" if disaster_type in ["concurrency", "payment"] else "medium"
})
return found
def _classify_disaster(self, keyword: str) -> str:
"""分類災難類型"""
disaster_map = {
"併發": "concurrency",
"並發": "concurrency",
"多用戶": "concurrency",
"同時": "concurrency",
"數據": "consistency",
"資料": "consistency",
"狀態": "consistency",
"交易": "consistency",
"支付": "payment",
"訂單": "payment",
"金錢": "payment",
"庫存": "payment",
"用戶": "security",
"認證": "security",
"權限": "security",
"密碼": "security",
"查詢": "performance",
"搜索": "performance",
"列表": "performance",
"分頁": "performance"
}
return disaster_map.get(keyword, "general")
def _calculate_complexity(self, requirement: str) -> int:
"""
計算複雜度分數 (0-10)
依據：
- 字數（越長越複雜）
- 功能數量（「、」的數量）
- 系統數量（「系統」、「平台」、「服務」的數量）
"""
word_count = len(requirement)
feature_count = requirement.count('、') + requirement.count('，')
system_count = (
requirement.count('系統') +
requirement.count('平台') +
requirement.count('服務')
)
# 計算分數
score = 0
if word_count > 100:
score += 3
elif word_count > 50:
score += 2
elif word_count > 20:
score += 1
score += min(feature_count, 4) # 最多加4分
score += min(system_count * 2, 3) # 最多加3分
return min(score, 10)
def _generate_summary(
self,
fuzzy_points: List[Dict],
disaster_scenarios: List[Dict],
complexity_score: int
) -> str:
"""生成分析摘要"""
parts = []
if len(fuzzy_points) > 0:
parts.append(f"發現 {len(fuzzy_points)} 個模糊描述")
if len(disaster_scenarios) > 0:
parts.append(f"發現 {len(disaster_scenarios)} 個潛在災難場景")
if complexity_score > 7:
parts.append("需求複雜度高")
elif complexity_score > 4:
parts.append("需求複雜度中等")
if not parts:
return "需求描述清晰，無明顯邏輯漏洞"
return "、".join(parts) + "，建議進行蘇格拉底面試"
# 全局單例
_analyzer = None
def get_analyzer() -> RequirementAnalyzer:
"""獲取分析器單例"""
global _analyzer
if _analyzer is None:
_analyzer = RequirementAnalyzer()
return _analyzer
if __name__ == "__main__":
# 測試分析器
analyzer = get_analyzer()
test_cases = [
"建立一個簡單的部落格系統",
"建立一個電商平台，需要處理高並發訂單、支付、庫存管理等功能",
"我想做一個類似 Uber 的叫車系統，用戶可以快速叫車，司機可以即時接單"
]
print("🔍 需求分析測試")
print("=" * 60)
for i, req in enumerate(test_cases, 1):
print(f"\n測試 {i}: {req[:30]}...")
result = analyzer.analyze(req)
print(f" 需要面試: {'✅ 是' if result['needs_interview'] else '❌ 否'}")
print(f" 模糊點: {len(result['fuzzy_points'])} 個")
print(f" 災難場景: {len(result['disaster_scenarios'])} 個")
print(f" 複雜度: {result['complexity_score']}/10")
print(f" 摘要: {result['analysis_summary']}")
