feat: 工资查询助手首版 v1.0.105

基于 skill-template 脚手架开发,包含:
- payroll_config(费率配置,可编辑)+ payslip_snapshots(解析历史,只读)2 张业务表
- 全国默认 + 北上广 4 套费率(社保/公积金/个税起征点/7 级税率表)
- 5 个业务 Action:calculate-payslip-simple / calculate-payslip-full / explain-term / view-config / edit-config
- 个税累计预扣法计算,支持多月历史累加
- 20 个薪资名词库(社保/公积金/个税/全勤/加班费等)
- 完整市场文档(README/TUTORIAL/DEMO/CHANGELOG)
- 296 测试全部通过
This commit is contained in:
skill-dev
2026-09-07 17:02:05 +08:00
commit 3db196e49d
205 changed files with 29927 additions and 0 deletions

View File

@@ -0,0 +1,592 @@
"""数据管理展示元数据校验(模板开发自检,不依赖宿主)。"""
from __future__ import annotations
import json
import os
import re
import sqlite3
from dataclasses import dataclass, field
from db.display_metadata import (
DATETIME_UNIX_SECONDS,
METADATA_TABLES,
STANDARD_TIMESTAMP_COLUMNS,
iter_user_tables,
list_editable_business_columns,
list_physical_column_names,
table_allows_manual_create,
)
from db.timestamp_columns import has_updated_at_trigger
from util.constants import SKILL_SLUG
from util.slug_naming import LEGACY_EXEMPT_SLUGS, validate_slug_semantics
_TEMPLATE_PLACEHOLDER_SLUGS = frozenset({"query-payslip-default", "query_payslip_default"})
_ENGLISH_PLACEHOLDER_NAME_PATTERNS = (
re.compile(r"^my\s+skill$", re.I),
re.compile(r"^your[-_\s]?skill", re.I),
re.compile(r"^skill[-_\s]?template$", re.I),
re.compile(r"^[a-z0-9_-]+$"),
)
_KEBAB_SLUG = re.compile(r"^[a-z0-9]+(?:-[a-z0-9]+)*$")
_CJK_RE = re.compile(r"[\u4e00-\u9fff]")
_FK_DDL_PATTERN = re.compile(r"\bFOREIGN\s+KEY\b|\bREFERENCES\b", re.I)
@dataclass
class ValidationReport:
errors: list[str] = field(default_factory=list)
warnings: list[str] = field(default_factory=list)
@property
def ok(self) -> bool:
return not self.errors
def _has_cjk(text: str) -> bool:
return bool(_CJK_RE.search(text))
def _looks_like_raw_physical_name(display_name: str, physical_name: str) -> bool:
if not display_name.strip():
return True
if display_name.strip().lower() == physical_name.strip().lower():
return True
compact = display_name.replace(" ", "").replace("_", "").lower()
if compact == physical_name.replace("_", "").lower():
return True
return False
def _parse_frontmatter_field(text: str, field_name: str) -> str:
parts = text.split("---", 2)
if len(parts) < 2:
return ""
for line in parts[1].splitlines():
match = re.match(rf"^{re.escape(field_name)}:\s*(.+)\s*$", line.strip())
if match:
return match.group(1).strip().strip("\"'")
return ""
def _parse_openclaw_slug(text: str) -> str:
parts = text.split("---", 2)
if len(parts) < 2:
return ""
lines = parts[1].splitlines()
for index, line in enumerate(lines):
if re.match(r"^\s+openclaw:\s*$", line):
for inner in lines[index + 1 :]:
if inner.strip() and inner[0] not in (" ", "\t"):
break
match = re.match(r"^(\s+)slug:\s*(.+)\s*$", inner)
if match:
return match.group(2).strip().strip("\"'")
return ""
def parse_skill_frontmatter(skill_root: str) -> dict[str, str]:
path = os.path.join(skill_root, "SKILL.md")
with open(path, encoding="utf-8") as f:
text = f.read()
return {
"name": _parse_frontmatter_field(text, "name"),
"slug": _parse_openclaw_slug(text),
}
def validate_skill_frontmatter(skill_root: str, *, allow_template_placeholders: bool = True) -> ValidationReport:
report = ValidationReport()
data = parse_skill_frontmatter(skill_root)
name = str(data.get("name") or "").strip()
slug = str(data.get("slug") or "").strip()
if not name:
report.errors.append("SKILL.md frontmatter.name 缺失")
elif not _has_cjk(name):
if allow_template_placeholders and slug in _TEMPLATE_PLACEHOLDER_SLUGS:
report.warnings.append("SKILL.md frontmatter.name 建议使用清晰的中文业务名称(模板占位可暂保留)")
else:
report.errors.append("SKILL.md frontmatter.name 应使用用户可见的中文业务名称")
else:
for pattern in _ENGLISH_PLACEHOLDER_NAME_PATTERNS:
if pattern.fullmatch(name):
if allow_template_placeholders and slug in _TEMPLATE_PLACEHOLDER_SLUGS:
report.warnings.append(f"SKILL.md frontmatter.name 仍为占位风格:{name!r}")
else:
report.errors.append(f"SKILL.md frontmatter.name 不应为英文占位名:{name!r}")
break
if not slug:
report.errors.append("SKILL.md metadata.openclaw.slug 缺失")
elif not _KEBAB_SLUG.fullmatch(slug):
report.errors.append(f"metadata.openclaw.slug 必须为 kebab-case:{slug!r}")
elif slug in _TEMPLATE_PLACEHOLDER_SLUGS and not allow_template_placeholders:
report.errors.append(f"metadata.openclaw.slug 仍为模板占位:{slug!r}")
constants_slug = SKILL_SLUG.strip()
if slug and constants_slug and slug != constants_slug:
report.warnings.append(
f"SKILL.md slug ({slug!r}) 与 util.constants.SKILL_SLUG ({constants_slug!r}) 不一致"
)
if slug and slug not in LEGACY_EXEMPT_SLUGS and _KEBAB_SLUG.fullmatch(slug):
naming_errors, naming_warnings = validate_slug_semantics(
slug,
strict=not allow_template_placeholders,
)
report.errors.extend(naming_errors)
report.warnings.extend(naming_warnings)
return report
def _quote_identifier(name: str) -> str:
return '"' + name.replace('"', '""') + '"'
def _table_column_info(cursor, table_name: str) -> dict[str, dict[str, object]]:
cursor.execute(f"PRAGMA table_info({_quote_identifier(table_name)})")
info: dict[str, dict[str, object]] = {}
for row in cursor.fetchall():
info[str(row[1])] = {
"cid": int(row[0]),
"name": str(row[1]),
"type": str(row[2] or ""),
"notnull": int(row[3] or 0),
"dflt_value": row[4],
"pk": int(row[5] or 0),
}
return info
def _table_column_xinfo(cursor, table_name: str) -> dict[str, dict[str, object]]:
"""PRAGMA table_xinfo 可识别 generated/hidden 字段(SQLite 3.31+)。"""
cursor.execute(f"PRAGMA table_xinfo({_quote_identifier(table_name)})")
info: dict[str, dict[str, object]] = {}
for row in cursor.fetchall():
hidden = int(row[6] or 0) if len(row) > 6 else 0
info[str(row[1])] = {
"cid": int(row[0]),
"name": str(row[1]),
"type": str(row[2] or ""),
"notnull": int(row[3] or 0),
"dflt_value": row[4],
"pk": int(row[5] or 0),
"hidden": hidden,
}
return info
def _is_generated_column(xinfo: dict[str, object]) -> bool:
hidden = int(xinfo.get("hidden") or 0)
return hidden in (2, 3)
def _is_blob_column(column_info: dict[str, object]) -> bool:
return "BLOB" in str(column_info.get("type") or "").upper()
def _default_uses_unixepoch(default_value: object) -> bool:
if default_value is None:
return False
text = str(default_value).strip().lower()
return "unixepoch" in text
def validate_options_json(options_json: str | None) -> tuple[list[str], list[str]]:
"""校验 options_json:静态枚举数组,或 lookup 对象。"""
errors: list[str] = []
warnings: list[str] = []
if options_json is None or str(options_json).strip() == "":
return errors, warnings
try:
parsed = json.loads(str(options_json))
except json.JSONDecodeError as exc:
errors.append(f"options_json 不是合法 JSON:{exc}")
return errors, warnings
if isinstance(parsed, dict):
for key in ("ref_table", "value_column", "label_column"):
if not str(parsed.get(key) or "").strip():
errors.append(f"options_json lookup 必须包含非空 {key}")
ident = re.compile(r"^[a-zA-Z_][a-zA-Z0-9_]*$")
for key in ("ref_table", "value_column", "label_column", "filter_column"):
raw = parsed.get(key)
if raw is None or raw == "":
continue
if not ident.match(str(raw)):
errors.append(f"options_json.{key} 必须是安全标识符,当前为 {raw!r}")
if parsed.get("filter_column") not in (None, "") and "filter_value" not in parsed:
warnings.append("options_json lookup 声明了 filter_column 但未提供 filter_value")
if "multiple" in parsed and not isinstance(parsed.get("multiple"), bool):
errors.append("options_json lookup.multiple 必须是布尔值")
return errors, warnings
if not isinstance(parsed, list):
errors.append("options_json 必须是 JSON 数组(静态枚举)或对象(lookup)")
return errors, warnings
for index, item in enumerate(parsed):
if not isinstance(item, dict):
errors.append(f"options_json[{index}] 必须是对象")
continue
if "value" not in item or "label" not in item:
errors.append(f"options_json[{index}] 必须包含 value 与 label")
return errors, warnings
def validate_filter_search_metadata(
cursor,
table_name: str,
) -> tuple[list[str], list[str]]:
"""对可见但未开启 searchable 的枚举 / boolean 列给出软建议。"""
warnings: list[str] = []
cursor.execute(
"""
SELECT column_name, visible, searchable, display_type, options_json
FROM _jiangchang_columns
WHERE table_name = ?
""",
(table_name,),
)
for column_name, visible, searchable, display_type, options_json in cursor.fetchall():
if visible is not None and int(visible) == 0:
continue
if searchable is not None and int(searchable) != 0:
continue
display_type_norm = str(display_type or "").strip()
is_boolean = display_type_norm == "boolean"
has_enum = False
if options_json is not None and str(options_json).strip():
opt_errors, _ = validate_options_json(options_json)
if not opt_errors:
try:
parsed = json.loads(str(options_json))
except json.JSONDecodeError:
parsed = None
has_enum = isinstance(parsed, list) and len(parsed) > 0
if not has_enum and not is_boolean:
continue
warnings.append(
f"{table_name}.{column_name}: 枚举/状态类列建议 searchable=1,"
"以便宿主列筛选与按中文 label 全局搜索(见 references/SCHEMA.md)"
)
return [], warnings
def validate_column_permissions(
cursor,
table_name: str,
*,
physical_columns: list[str] | None = None,
) -> tuple[list[str], list[str]]:
"""校验常见系统字段 editable=0,并识别技术上可写但不适合手工新增的表。"""
errors: list[str] = []
warnings: list[str] = []
ordered = physical_columns or list_physical_column_names(cursor, table_name)
column_info = _table_column_info(cursor, table_name)
column_xinfo = _table_column_xinfo(cursor, table_name)
cursor.execute(
"""
SELECT column_name, editable, options_json
FROM _jiangchang_columns
WHERE table_name = ?
""",
(table_name,),
)
meta_by_column = {
str(row[0]): {"editable": row[1], "options_json": row[2]}
for row in cursor.fetchall()
}
for column_name, meta in meta_by_column.items():
physical = column_info.get(column_name) or {}
xinfo = column_xinfo.get(column_name) or physical
if not physical and column_name not in column_xinfo:
continue
pk = int(physical.get("pk") or xinfo.get("pk") or 0)
editable = meta["editable"]
if pk and editable is not None and int(editable) != 0:
errors.append(f"{table_name}.{column_name} 是主键,editable 应为 0")
if column_name == "id" and editable is not None and int(editable) != 0:
errors.append(f"{table_name}.id 的 editable 应为 0(自增主键)")
if _is_generated_column(xinfo) and editable is not None and int(editable) != 0:
errors.append(f"{table_name}.{column_name} 是 generated 字段,editable 应为 0")
if _is_blob_column(physical or xinfo) and editable is not None and int(editable) != 0:
errors.append(f"{table_name}.{column_name} 是 BLOB 字段,editable 应为 0")
opt_errors, _ = validate_options_json(meta.get("options_json"))
for item in opt_errors:
errors.append(f"{table_name}.{column_name}: {item}")
cursor.execute(
"SELECT readonly FROM _jiangchang_tables WHERE table_name = ?",
(table_name,),
)
readonly_row = cursor.fetchone()
readonly = int(readonly_row[0]) if readonly_row else None
editable_columns = list_editable_business_columns(cursor, table_name)
if readonly == 0 and not editable_columns:
warnings.append(
f"{table_name}: 表级 readonly=0 但所有字段 editable=0,"
"宿主不应打开空白新增表单;数据应由技能业务流程产生"
)
if readonly == 1 and editable_columns:
warnings.append(
f"{table_name}: 表级 readonly=1 但存在 editable=1 字段,"
"宿主应以表级只读为准"
)
return errors, warnings
def _expected_timestamp_display_name(column_name: str) -> str:
if column_name == "created_at":
return "创建时间"
if column_name == "updated_at":
return "更新时间"
return ""
def validate_standard_timestamp_columns(
cursor,
table_name: str,
*,
physical_columns: list[str] | None = None,
) -> tuple[list[str], list[str]]:
errors: list[str] = []
warnings: list[str] = []
ordered = physical_columns or list_physical_column_names(cursor, table_name)
column_info = _table_column_info(cursor, table_name)
for column_name in STANDARD_TIMESTAMP_COLUMNS:
if column_name not in ordered:
continue
physical = column_info.get(column_name)
if not physical:
errors.append(f"{table_name}.{column_name} 缺少物理字段定义")
continue
if "INT" not in str(physical["type"]).upper():
errors.append(
f"{table_name}.{column_name} 字段类型应为 INTEGER(Unix 秒级时间戳),当前为 {physical['type']!r}"
)
if column_name == "created_at" and not _default_uses_unixepoch(physical["dflt_value"]):
errors.append(
f"{table_name}.created_at 应设置数据库默认值 DEFAULT (unixepoch()),当前默认值为 {physical['dflt_value']!r}"
)
if column_name == "updated_at" and not _default_uses_unixepoch(physical["dflt_value"]):
errors.append(
f"{table_name}.updated_at 应设置数据库初始默认值 DEFAULT (unixepoch()),当前默认值为 {physical['dflt_value']!r}"
)
cursor.execute(
"""
SELECT display_name, display_type, editable, searchable, display_order
FROM _jiangchang_columns
WHERE table_name = ? AND column_name = ?
""",
(table_name, column_name),
)
meta_row = cursor.fetchone()
if not meta_row:
errors.append(f"{table_name}.{column_name} 缺少 _jiangchang_columns 元数据")
continue
display_name, display_type, editable, searchable, display_order = meta_row
expected_name = _expected_timestamp_display_name(column_name)
if str(display_name or "").strip() != expected_name:
errors.append(
f"{table_name}.{column_name} 的 display_name 应为 {expected_name!r},当前为 {display_name!r}"
)
if str(display_type or "").strip() != DATETIME_UNIX_SECONDS:
errors.append(
f"{table_name}.{column_name} 的 display_type 应为 {DATETIME_UNIX_SECONDS!r},当前为 {display_type!r}"
)
if editable is None or int(editable) != 0:
errors.append(f"{table_name}.{column_name} 的 editable 应为 0(系统时间字段不可编辑)")
if display_order is not None:
errors.append(f"{table_name}.{column_name} 的 display_order 应为 NULL")
if searchable is not None and int(searchable) != 0:
warnings.append(f"{table_name}.{column_name} 的时间字段通常不必开启 searchable")
if "created_at" in ordered and "updated_at" in ordered:
if ordered.index("updated_at") < ordered.index("created_at"):
warnings.append(f"{table_name}: updated_at 不应出现在 created_at 之前")
if "updated_at" in ordered and not has_updated_at_trigger(cursor, table_name):
errors.append(
f"{table_name}.updated_at 缺少自动维护 trigger(期望 {table_name}_set_updated_at)"
)
return errors, warnings
def column_order_warnings(table_name: str, ordered_names: list[str]) -> list[str]:
warnings: list[str] = []
if "id" in ordered_names and ordered_names.index("id") != 0:
warnings.append(f"{table_name}: 字段 id 通常应定义在 CREATE TABLE 的第一列")
if "created_at" in ordered_names:
idx = ordered_names.index("created_at")
trailing = ordered_names[idx + 1 :]
if trailing and trailing != ["updated_at"]:
warnings.append(
f"{table_name}: created_at 之后通常只应跟 updated_at,不应再放置其他业务字段"
)
if "updated_at" in ordered_names and "created_at" in ordered_names:
if ordered_names.index("updated_at") < ordered_names.index("created_at"):
warnings.append(f"{table_name}: updated_at 不应出现在 created_at 之前")
return warnings
def _table_exists(cursor, table_name: str) -> bool:
cursor.execute(
"SELECT 1 FROM sqlite_master WHERE type = 'table' AND name = ? LIMIT 1",
(table_name,),
)
return cursor.fetchone() is not None
def validate_no_sqlite_foreign_keys(cursor) -> tuple[list[str], list[str]]:
"""业务表 DDL 不得含 FOREIGN KEY / REFERENCES(关联走 repository/service)。"""
errors: list[str] = []
cursor.execute(
"""
SELECT name, sql FROM sqlite_master
WHERE type = 'table' AND name NOT LIKE 'sqlite_%'
"""
)
for table_name, sql in cursor.fetchall():
if table_name in METADATA_TABLES:
continue
ddl = str(sql or "")
if ddl and _FK_DDL_PATTERN.search(ddl):
errors.append(
f"业务表 {table_name!r} 的 DDL 含 SQLite FOREIGN KEY/REFERENCES;"
"请改为业务 id 列 + repository/service 逻辑关联(见 references/SCHEMA.md)"
)
return errors, []
def validate_display_metadata(conn: sqlite3.Connection, *, check_column_order: bool = True) -> ValidationReport:
report = ValidationReport()
cur = conn.cursor()
has_tables_meta = _table_exists(cur, "_jiangchang_tables")
has_columns_meta = _table_exists(cur, "_jiangchang_columns")
if not has_tables_meta:
report.errors.append("缺少数据管理元数据表 _jiangchang_tables")
if not has_columns_meta:
report.errors.append("缺少数据管理元数据表 _jiangchang_columns")
if not has_tables_meta or not has_columns_meta:
return report
fk_errors, _ = validate_no_sqlite_foreign_keys(cur)
report.errors.extend(fk_errors)
for table_name in iter_user_tables(cur):
cur.execute(
"SELECT display_name FROM _jiangchang_tables WHERE table_name = ?",
(table_name,),
)
row = cur.fetchone()
if not row or not str(row[0] or "").strip():
report.errors.append(f"业务表 {table_name!r} 缺少 _jiangchang_tables.display_name")
elif not _has_cjk(str(row[0])):
report.errors.append(f"业务表 {table_name!r} 的 display_name 应使用中文业务名称")
physical_columns = list_physical_column_names(cur, table_name)
if check_column_order:
report.warnings.extend(column_order_warnings(table_name, physical_columns))
for column_name in physical_columns:
cur.execute(
"""
SELECT display_name, visible
FROM _jiangchang_columns
WHERE table_name = ? AND column_name = ?
""",
(table_name, column_name),
)
col_row = cur.fetchone()
if not col_row:
report.errors.append(
f"字段 {table_name}.{column_name} 缺少 _jiangchang_columns 元数据"
)
continue
display_name = str(col_row[0] or "").strip()
visible = col_row[1] is None or int(col_row[1]) != 0
if visible and not display_name:
report.errors.append(f"字段 {table_name}.{column_name} 的 display_name 为空")
elif visible and _looks_like_raw_physical_name(display_name, column_name):
report.errors.append(
f"字段 {table_name}.{column_name} 的 display_name 不应直接等于英文物理名:{display_name!r}"
)
elif visible and not _has_cjk(display_name):
report.errors.append(
f"字段 {table_name}.{column_name} 的 display_name 应使用中文:{display_name!r}"
)
ts_errors, ts_warnings = validate_standard_timestamp_columns(
cur,
table_name,
physical_columns=physical_columns,
)
report.errors.extend(ts_errors)
report.warnings.extend(ts_warnings)
perm_errors, perm_warnings = validate_column_permissions(
cur,
table_name,
physical_columns=physical_columns,
)
report.errors.extend(perm_errors)
report.warnings.extend(perm_warnings)
_, filter_warnings = validate_filter_search_metadata(cur, table_name)
report.warnings.extend(filter_warnings)
cur.execute("SELECT table_name, column_name FROM _jiangchang_columns")
for table_name, column_name in cur.fetchall():
if table_name in METADATA_TABLES:
continue
cur.execute(
"SELECT 1 FROM sqlite_master WHERE type='table' AND name = ?",
(table_name,),
)
if not cur.fetchone():
report.errors.append(f"元数据引用了不存在的表:{table_name!r}")
continue
physical = list_physical_column_names(cur, str(table_name))
if str(column_name) not in physical:
report.errors.append(f"元数据引用了不存在的字段:{table_name}.{column_name}")
cur.execute("SELECT table_name FROM _jiangchang_tables")
for (table_name,) in cur.fetchall():
if table_name in METADATA_TABLES:
continue
cur.execute(
"SELECT 1 FROM sqlite_master WHERE type='table' AND name = ?",
(table_name,),
)
if not cur.fetchone():
report.errors.append(f"元数据引用了不存在的表:{table_name!r}")
return report
def validate_template_database(conn: sqlite3.Connection) -> ValidationReport:
"""模板自检入口:校验展示元数据完整性,并扫描业务表 DDL 是否含 SQLite 外键。"""
return validate_display_metadata(conn)