API 版本与破坏性变更检测引擎（专业版）

输入

detect_breaking(old,new)；is_breaking_change(old,new)；semver_bump(breaking,added_optional,patch_only)
字段值可写类型字符串，或 {{"type":"string","required":True}}

输出

破坏性变更元组列表 (类型, 详情)；是否 breaking；major/minor/patch

硬性规则

1. 破坏性四类：field_removed/type_changed/became_required/new_required_field。
2. 有任一破坏性→major；仅新增可选→minor；否则 patch。
3. 字段缺失与类型变化分开报告，不合并。

示例

detect_breaking({{"id":"string"}},{{}}) → [("field_removed","id")]
is_breaking_change({{"a":"string"}},{{"a":"integer"}}) → True
semver_bump(False, added_optional=True) → "minor"

Python 引擎（可独立运行）

# -*- coding: utf-8 -*-
def detect_breaking(old, new):
    """old/new: {字段: 类型或 {'type':..,'required':bool}}；返回破坏性变更列表"""
    def norm(f):
        if isinstance(f, dict): return f.get("type"), bool(f.get("required", False))
        return f, False
    breaks = []
    for name, spec in old.items():
        ot, oreq = norm(spec)
        if name not in new:
            breaks.append(("field_removed", name)); continue
        nt, nreq = norm(new[name])
        if ot != nt:
            breaks.append(("type_changed", f"{name}:{ot}->{nt}"))
        if (not oreq) and nreq:
            breaks.append(("became_required", name))
    for name, spec in new.items():
        if name not in old and norm(spec)[1]:
            breaks.append(("new_required_field", name))
    return breaks

def semver_bump(breaking, added_optional=False, patch_only=False):
    if breaking: return "major"
    if added_optional: return "minor"
    if patch_only: return "patch"
    return "patch"

def is_breaking_change(old, new):
    return len(detect_breaking(old, new)) > 0