aukeys-opscli 0.0.6__py3-none-any.whl

opscli/skills/templates/ops-dataset-query/scripts/core.py

@@ -0,0 +1,140 @@
+"""ops-dataset-query Skill 的核心工具函数。
+
+提供 CSV 加载、过滤和搜索打分能力，供 search.py 和 updater.py 复用。
+"""
+
+from __future__ import annotations
+
+import csv
+import re
+from pathlib import Path
+
+
+def load_csv_rows(path: Path) -> list[dict]:
+    """加载 CSV 文件为字典列表。
+
+    使用 utf-8-sig 编码以兼容带 BOM 的 CSV 文件。
+    文件不存在时返回空列表。
+
+    Args:
+        path: CSV 文件路径
+
+    Returns:
+        每行一个字典，键名为 CSV 表头
+    """
+    if not path.exists():
+        return []
+    with path.open("r", encoding="utf-8-sig", newline="") as handle:
+        return list(csv.DictReader(handle))
+
+
+def filter_rows_by_dataset(rows: list[dict], dataset: str | None) -> list[dict]:
+    """按数据集别名过滤行。
+
+    Args:
+        rows: 原始字段行
+        dataset: 数据集别名；为空时不过滤
+
+    Returns:
+        过滤后的字段行
+    """
+    if not dataset:
+        return rows
+
+    normalized = dataset.strip().lower()
+    if not normalized:
+        return rows
+
+    return [
+        row for row in rows
+        if str(row.get("dataset_alias", "")).strip().lower() == normalized
+    ]
+
+
+def search_rows(rows: list[dict], keyword: str, limit: int = 10) -> list[dict]:
+    """按关键词搜索字段行，并返回按相关性排序后的结果。
+
+    排序策略使用简单加权打分，优先匹配：
+    - `field_name`
+    - `verbose_name`
+    - `keywords`
+    - `description`
+    - 其他整行内容
+    """
+    normalized = keyword.strip().lower()
+    if not normalized:
+        return []
+
+    tokens = _tokenize(normalized)
+    if not tokens:
+        return []
+
+    scored: list[tuple[int, dict]] = []
+    for row in rows:
+        score = _score_row(row, normalized, tokens)
+        if score <= 0:
+            continue
+        scored.append((score, row))
+
+    scored.sort(
+        key=lambda item: (
+            -item[0],
+            str(item[1].get("dataset_alias", "")),
+            str(item[1].get("field_name", "")),
+        )
+    )
+
+    return [row for _, row in scored[: max(limit, 0)]]
+
+
+def _tokenize(value: str) -> list[str]:
+    """将搜索词切分为 token。"""
+    return [token for token in re.split(r"[\s_\-./]+", value.lower()) if token]
+
+
+def _score_row(row: dict, keyword: str, tokens: list[str]) -> int:
+    """为单行结果打分。"""
+    field_name = str(row.get("field_name", "")).lower()
+    verbose_name = str(row.get("verbose_name", "")).lower()
+    keywords = str(row.get("keywords", "")).lower()
+    description = str(row.get("description", "")).lower()
+    dataset_alias = str(row.get("dataset_alias", "")).lower()
+    dataset_name = str(row.get("dataset_name", "")).lower()
+    fulltext = " ".join(str(value).lower() for value in row.values())
+
+    score = 0
+
+    # 精确/短字段优先
+    if keyword == field_name:
+        score += 120
+    if keyword == verbose_name:
+        score += 100
+    if keyword == dataset_alias:
+        score += 40
+
+    # 连续子串匹配
+    if keyword in field_name:
+        score += 60
+    if keyword in verbose_name:
+        score += 45
+    if keyword in keywords:
+        score += 35
+    if keyword in dataset_name:
+        score += 20
+    if keyword in description:
+        score += 10
+
+    # token 逐项匹配
+    score += _token_match_score(field_name, tokens, 16)
+    score += _token_match_score(verbose_name, tokens, 12)
+    score += _token_match_score(keywords, tokens, 9)
+    score += _token_match_score(dataset_name, tokens, 6)
+    score += _token_match_score(description, tokens, 4)
+    score += _token_match_score(fulltext, tokens, 2)
+
+    return score
+
+
+def _token_match_score(text: str, tokens: list[str], weight: int) -> int:
+    """计算 token 命中的加权分数。"""
+    return sum(weight for token in tokens if token in text)

opscli/skills/templates/ops-dataset-query/scripts/query.py

@@ -0,0 +1,145 @@
+"""ops-dataset-query Skill 的查询转发脚本。
+
+该脚本不直接访问后端接口，只负责把查询动作转交给 `opscli query`。
+适合在 Skill 目录内直接运行：
+
+- `python query.py metadata --dataset ds_xxx`
+- `python query.py run --payload payload.json`
+"""
+
+from __future__ import annotations
+
+import argparse
+import importlib.util
+import json
+import os
+import shutil
+import subprocess
+import sys
+
+
+def build_opscli_prefix() -> list[str]:
+    """解析 opscli 调用前缀。
+
+    优先级：
+    1. 环境变量 OPSCLI_BIN
+    2. PATH 中的 opscli 可执行文件
+    3. 当前 Python 环境可导入 opscli.cli 时使用 `python -m opscli.cli`
+    """
+    env_bin = os.getenv("OPSCLI_BIN")
+    if env_bin:
+        return [env_bin]
+
+    cli_bin = shutil.which("opscli")
+    if cli_bin:
+        return [cli_bin]
+
+    if importlib.util.find_spec("opscli.cli") is not None:
+        return [sys.executable, "-m", "opscli.cli"]
+
+    raise RuntimeError("未找到 opscli，请先安装 opscli，或通过 OPSCLI_BIN 指定可执行路径")
+
+
+def build_command(args: argparse.Namespace) -> list[str]:
+    """根据脚本参数构造最终的 opscli query 命令。"""
+    command = [*build_opscli_prefix(), "query", args.command]
+
+    if args.command == "metadata":
+        if args.dataset:
+            command.extend(["--dataset", args.dataset])
+        if args.table_id is not None:
+            command.extend(["--table-id", str(args.table_id)])
+        if args.skills_dir:
+            command.extend(["--skills-dir", args.skills_dir])
+    elif args.command == "run":
+        command.extend(["--payload", args.payload])
+    elif args.command == "build":
+        if args.dataset:
+            command.extend(["--dataset", args.dataset])
+        if args.table_id is not None:
+            command.extend(["--table-id", str(args.table_id)])
+        for item in args.dimension or []:
+            command.extend(["--dimension", item])
+        for item in args.metric or []:
+            command.extend(["--metric", item])
+        for item in args.where or []:
+            command.extend(["--where", item])
+        for item in args.order_by or []:
+            command.extend(["--order-by", item])
+        if args.where_json:
+            command.extend(["--where-json", args.where_json])
+        if args.where_file:
+            command.extend(["--where-file", args.where_file])
+        if args.skills_dir:
+            command.extend(["--skills-dir", args.skills_dir])
+        if args.output:
+            command.extend(["--output", args.output])
+        command.extend(["--limit", str(args.limit), "--offset", str(args.offset)])
+        if args.run:
+            command.append("--run")
+
+    if args.pretty:
+        command.append("--pretty")
+    return command
+
+
+def _emit_error(message: str) -> None:
+    """以 JSON 形式输出本地错误，保持脚本行为稳定。"""
+    print(
+        json.dumps(
+            {
+                "success": False,
+                "command": "ops-dataset-query query",
+                "data": None,
+                "error": message,
+            },
+            ensure_ascii=False,
+        )
+    )
+
+
+def main() -> None:
+    """运行查询转发脚本。"""
+    parser = argparse.ArgumentParser()
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    metadata = subparsers.add_parser("metadata")
+    metadata.add_argument("--dataset", help="dataset_alias")
+    metadata.add_argument("--table-id", type=int, help="table_id")
+    metadata.add_argument("--skills-dir", help="指定 Skill 目录")
+    metadata.add_argument("--pretty", action="store_true", help="格式化输出")
+
+    run = subparsers.add_parser("run")
+    run.add_argument("--payload", required=True, help="查询 payload JSON 文件")
+    run.add_argument("--pretty", action="store_true", help="格式化输出")
+
+    build = subparsers.add_parser("build")
+    build.add_argument("--dataset", help="dataset_alias")
+    build.add_argument("--table-id", type=int, help="table_id")
+    build.add_argument("--dimension", action="append", help="维度定义：field_name[:alias]")
+    build.add_argument("--metric", action="append", help="指标定义：field_name:aggregation[:alias]")
+    build.add_argument("--where", action="append", help="筛选条件：field|operator|value_json，可重复")
+    build.add_argument("--where-json", help="where JSON 字符串")
+    build.add_argument("--where-file", help="where JSON 文件路径")
+    build.add_argument("--order-by", action="append", help="排序定义：expr[:asc|desc]")
+    build.add_argument("--limit", type=int, default=20, help="limit，默认 20")
+    build.add_argument("--offset", type=int, default=0, help="offset，默认 0")
+    build.add_argument("--skills-dir", help="指定 Skill 目录")
+    build.add_argument("--output", help="将 payload 写入指定文件")
+    build.add_argument("--run", action="store_true", help="构造后立即执行查询")
+    build.add_argument("--pretty", action="store_true", help="格式化输出")
+
+    args = parser.parse_args()
+
+    try:
+        command = build_command(args)
+    except Exception as exc:
+        _emit_error(str(exc))
+        raise SystemExit(1) from exc
+
+    result = subprocess.run(command, check=False)
+    raise SystemExit(result.returncode)
+
+
+if __name__ == "__main__":
+    main()

opscli/skills/templates/ops-dataset-query/scripts/search.py

@@ -0,0 +1,36 @@
+"""ops-dataset-query Skill 的搜索脚本。
+
+提供命令行关键词搜索功能，支持：
+- 关键词搜索
+- 指定数据集过滤
+- 限制返回数量
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+from core import filter_rows_by_dataset, load_csv_rows, search_rows
+
+
+def main() -> None:
+    """读取 CSV 数据并按关键词搜索，输出匹配结果。"""
+    parser = argparse.ArgumentParser()
+    parser.add_argument("keyword")
+    parser.add_argument("--dataset", help="按 dataset_alias 过滤")
+    parser.add_argument("-n", "--limit", type=int, default=10, help="限制返回条数，默认 10")
+    args = parser.parse_args()
+
+    # 数据目录位于脚本上级目录的 data/ 子目录
+    data_dir = Path(__file__).resolve().parent.parent / "data"
+    rows = load_csv_rows(data_dir / "dataset_fields.csv")
+    filtered_rows = filter_rows_by_dataset(rows, args.dataset)
+    matched_rows = search_rows(filtered_rows, args.keyword, limit=args.limit)
+
+    print(json.dumps(matched_rows, ensure_ascii=False, indent=2))
+
+
+if __name__ == "__main__":
+    main()

opscli/skills/templates/ops-dataset-query/scripts/updater.py

@@ -0,0 +1,106 @@
+"""ops-dataset-query Skill 的本地更新脚本。
+
+提供独立于 `opscli skills upgrade` 的本地入口，适合在 Skill 目录内直接运行：
+
+- `python updater.py --check`：仅检查远端版本
+- `python updater.py`：执行更新
+- `python updater.py --force`：强制重新拉取
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+from opscli.skills.models import SkillRecord
+from opscli.skills.updater import SkillsUpdater
+
+
+def build_record(skill_root: Path) -> SkillRecord:
+    """基于当前 Skill 目录构造记录对象。"""
+    data_dir = skill_root / "data"
+    version_file = data_dir / "VERSION.json"
+    version = "v0.0.0"
+    if version_file.exists():
+        payload = json.loads(version_file.read_text(encoding="utf-8"))
+        version = str(payload.get("version", "v0.0.0"))
+
+    return SkillRecord(
+        name="ops-dataset-query",
+        version=version,
+        runtime="custom",
+        root=skill_root,
+        version_file=version_file,
+    )
+
+
+def check_update(updater: SkillsUpdater, record: SkillRecord) -> dict:
+    """检查当前 Skill 是否存在远端更新。"""
+    manifest = updater.fetch_manifest(record.name)
+    if not manifest:
+        raise ValueError("远端 manifest 不存在")
+
+    remote_version = str(manifest.get("version", "v0.0.0"))
+    compare_result = updater.compare_versions(record.version, remote_version)
+
+    return {
+        "name": record.name,
+        "current_version": record.version,
+        "remote_version": remote_version,
+        "update_available": compare_result < 0,
+    }
+
+
+def _emit(payload: dict, pretty: bool) -> None:
+    """统一输出 JSON。"""
+    if pretty:
+        print(json.dumps(payload, ensure_ascii=False, indent=2))
+    else:
+        print(json.dumps(payload, ensure_ascii=False))
+
+
+def main() -> None:
+    """运行本地更新脚本。"""
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--check", action="store_true", help="仅检查远端版本，不执行更新")
+    parser.add_argument("--force", action="store_true", help="强制重新下载远端数据")
+    parser.add_argument("--pretty", action="store_true", help="格式化输出 JSON")
+    args = parser.parse_args()
+
+    skill_root = Path(__file__).resolve().parent.parent
+    updater = SkillsUpdater()
+    record = build_record(skill_root)
+
+    try:
+        if args.check:
+            payload = {
+                "success": True,
+                "command": "ops-dataset-query updater check",
+                "data": check_update(updater, record),
+                "error": None,
+            }
+        else:
+            result = updater.upgrade_ops_dataset_query(record, force=args.force)
+            payload = {
+                "success": True,
+                "command": "ops-dataset-query updater upgrade",
+                "data": result.to_dict(),
+                "error": None,
+            }
+        _emit(payload, args.pretty)
+    except Exception as exc:
+        _emit(
+            {
+                "success": False,
+                "command": "ops-dataset-query updater",
+                "data": None,
+                "error": str(exc),
+            },
+            args.pretty,
+        )
+        raise SystemExit(1) from exc
+
+
+if __name__ == "__main__":
+    main()