with one click
regex-vs-llm-structured-text
选择在解析结构化文本时使用正则表达式还是大型语言模型的决策框架——从正则表达式开始,仅在低置信度的边缘情况下添加大型语言模型。
Menu
选择在解析结构化文本时使用正则表达式还是大型语言模型的决策框架——从正则表达式开始,仅在低置信度的边缘情况下添加大型语言模型。
Create reproducible, cross-platform (macOS/Linux) development environments with Flox, a declarative Nix-based environment manager. Use when setting up project toolchains for any language, installing system-level dependencies (compilers, databases, native libs like openssl/BLAS), pinning exact package versions for a team, running local services (PostgreSQL, Redis, Kafka), onboarding developers with one command, or solving 'works on my machine' problems — including agent/vibe-coding setups that need project-scoped tools without sudo. Also use when the user mentions .flox/, manifest.toml, flox activate, or FloxHub.
Commercial-grade Python installer expert for Windows: Nuitka extreme compilation, dist slimming, DLL footprint analysis, and Inno Setup packaging to ship the smallest, fastest installers. Use only for advanced packaging/optimization (minimal size, fast startup), not basic script-to-exe conversion. 中文触发:Nuitka 极限优化、Python 商业打包、极限编译 Python、dist 瘦身、DLL 分析、最小安装包、最快启动、商业级打包风格
Use when a brand needs to discover or articulate its identity through structured multi-session interviews. Covers purpose, positioning, audience, personality, voice, narrative, and founder-brand tension across 8 modules using laddering, 5 Whys, and projective techniques. Produces a resumable session with disk-persisted state and a master brandbook (90_SYNTHESIS.md).
Use when a brand needs to discover or articulate its identity through structured multi-session interviews. Covers purpose, positioning, audience, personality, voice, narrative, and founder-brand tension across 8 modules using laddering, 5 Whys, and projective techniques. Produces a resumable session with disk-persisted state and a master brandbook (90_SYNTHESIS.md).
Use this skill to automate visual testing and UI interaction verification using browser automation after deploying features.
Visualize whether skills, rules, and agent definitions are actually followed — auto-generates scenarios at 3 prompt strictness levels, runs agents, classifies behavioral sequences, and reports compliance rates with full tool call timelines
| name | regex-vs-llm-structured-text |
| description | 选择在解析结构化文本时使用正则表达式还是大型语言模型的决策框架——从正则表达式开始,仅在低置信度的边缘情况下添加大型语言模型。 |
| origin | ECC |
一个用于解析结构化文本(测验、表单、发票、文档)的实用决策框架。核心见解是:正则表达式能以低成本、确定性的方式处理 95-98% 的情况。将昂贵的 LLM 调用留给剩余的边缘情况。
文本格式是否一致且重复?
├── 是 (>90% 遵循某种模式) → 从正则表达式开始
│ ├── 正则表达式处理 95%+ → 完成,无需 LLM
│ └── 正则表达式处理 <95% → 仅为边缘情况添加 LLM
└── 否 (自由格式,高度可变) → 直接使用 LLM
[正则表达式解析器] ─── 提取结构(95-98% 准确率)
│
▼
[文本清理器] ─── 去除噪声(标记、页码、伪影)
│
▼
[置信度评分器] ─── 标记低置信度提取项
│
├── 高置信度(≥0.95)→ 直接输出
│
└── 低置信度(<0.95)→ [LLM 验证器] → 输出
import re
from dataclasses import dataclass
@dataclass(frozen=True)
class ParsedItem:
id: str
text: str
choices: tuple[str, ...]
answer: str
confidence: float = 1.0
def parse_structured_text(content: str) -> list[ParsedItem]:
"""Parse structured text using regex patterns."""
pattern = re.compile(
r"(?P<id>\d+)\.\s*(?P<text>.+?)\n"
r"(?P<choices>(?:[A-D]\..+?\n)+)"
r"Answer:\s*(?P<answer>[A-D])",
re.MULTILINE | re.DOTALL,
)
items = []
for match in pattern.finditer(content):
choices = tuple(
c.strip() for c in re.findall(r"[A-D]\.\s*(.+)", match.group("choices"))
)
items.append(ParsedItem(
id=match.group("id"),
text=match.group("text").strip(),
choices=choices,
answer=match.group("answer"),
))
return items
标记可能需要 LLM 审核的项:
@dataclass(frozen=True)
class ConfidenceFlag:
item_id: str
score: float
reasons: tuple[str, ...]
def score_confidence(item: ParsedItem) -> ConfidenceFlag:
"""Score extraction confidence and flag issues."""
reasons = []
score = 1.0
if len(item.choices) < 3:
reasons.append("few_choices")
score -= 0.3
if not item.answer:
reasons.append("missing_answer")
score -= 0.5
if len(item.text) < 10:
reasons.append("short_text")
score -= 0.2
return ConfidenceFlag(
item_id=item.id,
score=max(0.0, score),
reasons=tuple(reasons),
)
def identify_low_confidence(
items: list[ParsedItem],
threshold: float = 0.95,
) -> list[ConfidenceFlag]:
"""Return items below confidence threshold."""
flags = [score_confidence(item) for item in items]
return [f for f in flags if f.score < threshold]
def validate_with_llm(
item: ParsedItem,
original_text: str,
client,
) -> ParsedItem:
"""Use LLM to fix low-confidence extractions."""
response = client.messages.create(
model="claude-haiku-4-5-20251001", # Cheapest model for validation
max_tokens=500,
messages=[{
"role": "user",
"content": (
f"Extract the question, choices, and answer from this text.\n\n"
f"Text: {original_text}\n\n"
f"Current extraction: {item}\n\n"
f"Return corrected JSON if needed, or 'CORRECT' if accurate."
),
}],
)
# Parse LLM response and return corrected item...
return corrected_item
def process_document(
content: str,
*,
llm_client=None,
confidence_threshold: float = 0.95,
) -> list[ParsedItem]:
"""Full pipeline: regex -> confidence check -> LLM for edge cases."""
# Step 1: Regex extraction (handles 95-98%)
items = parse_structured_text(content)
# Step 2: Confidence scoring
low_confidence = identify_low_confidence(items, confidence_threshold)
if not low_confidence or llm_client is None:
return items
# Step 3: LLM validation (only for flagged items)
low_conf_ids = {f.item_id for f in low_confidence}
result = []
for item in items:
if item.id in low_conf_ids:
result.append(validate_with_llm(item, content, llm_client))
else:
result.append(item)
return result
来自一个生产中的测验解析管道(410 个项目):
| 指标 | 值 |
|---|---|
| 正则表达式成功率 | 98.0% |
| 低置信度项目 | 8 (2.0%) |
| 所需 LLM 调用次数 | ~5 |
| 相比全 LLM 的成本节省 | ~95% |
| 测试覆盖率 | 93% |