@9000ai/cli 0.1.0

package/skills/video-transcription/scripts/video_transcription_api.py

@@ -0,0 +1,183 @@
+from __future__ import annotations
+
+import argparse
+import json
+import sys
+from pathlib import Path
+from typing import Any
+
+import requests
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+HUB_ROOT = REPO_ROOT / "9000AI-hub-9000AI"
+if str(HUB_ROOT) not in sys.path:
+    sys.path.insert(0, str(HUB_ROOT))
+
+from shared.runner import (  # noqa: E402
+    ModuleSpec,
+    configure_stdout_encoding,
+    load_json_file,
+    print_json,
+    request_json,
+    resolve_api_key,
+    resolve_base_url,
+    save_local_config,
+)
+
+SKILL_ROOT = Path(__file__).resolve().parents[1]
+MODULE_SPEC = ModuleSpec(
+    module="video-transcription",
+    skill_root=SKILL_ROOT,
+)
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="Video transcription API client")
+    parser.add_argument("--base-url", default=None)
+    parser.add_argument("--api-key", default=None)
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    subparsers.add_parser("configure", help="写入本地配置文件")
+    subparsers.add_parser("whoami", help="查看当前 key 对应的调用方")
+    subparsers.add_parser("capabilities", help="查看当前 key 已开通的能力")
+
+    submit = subparsers.add_parser("submit", help="提交批量视频转文字任务")
+    submit.add_argument("--json-file", required=True)
+
+    task = subparsers.add_parser("task", help="查询任务状态")
+    task.add_argument("--task-id", required=True)
+
+    results = subparsers.add_parser("results", help="查询任务结果")
+    results.add_argument("--task-id", required=True)
+
+    text = subparsers.add_parser("text", help="只提取转写原文文案")
+    text.add_argument("--task-id", required=True)
+
+    return parser
+
+
+def _normalize_output(output: object) -> dict[str, Any]:
+    if isinstance(output, dict):
+        return output
+    if isinstance(output, str):
+        try:
+            parsed = json.loads(output)
+            if isinstance(parsed, dict):
+                return parsed
+        except json.JSONDecodeError:
+            return {}
+    return {}
+
+
+def _fetch_transcription_payload(json_url: str) -> dict[str, Any]:
+    response = requests.get(json_url, timeout=60)
+    response.raise_for_status()
+    payload = response.json()
+    if not isinstance(payload, dict):
+        raise SystemExit("转写 JSON 返回格式不是 object")
+    return payload
+
+
+def main() -> None:
+    configure_stdout_encoding()
+    parser = build_parser()
+    args = parser.parse_args()
+
+    if args.command == "configure":
+        if not args.base_url or not args.api_key:
+            raise SystemExit("configure 需要同时传 --base-url 和 --api-key")
+        save_local_config(MODULE_SPEC, base_url=args.base_url, api_key=args.api_key)
+        print_json({"message": "配置已写入", "config_path": str(MODULE_SPEC.config_path)})
+        return
+
+    base_url = resolve_base_url(MODULE_SPEC, args.base_url)
+    api_key = resolve_api_key(MODULE_SPEC, args.api_key)
+
+    if args.command == "whoami":
+        print_json(
+            request_json(MODULE_SPEC, method="GET", base_url=base_url, api_key=api_key, path="/api/v1/auth/me")
+        )
+        return
+
+    if args.command == "capabilities":
+        print_json(
+            request_json(
+                MODULE_SPEC,
+                method="GET",
+                base_url=base_url,
+                api_key=api_key,
+                path="/api/v1/auth/capability-permissions",
+            )
+        )
+        return
+
+    if args.command == "submit":
+        print_json(
+            request_json(
+                MODULE_SPEC,
+                method="POST",
+                base_url=base_url,
+                api_key=api_key,
+                path="/api/v1/media/batch-video-to-text",
+                payload=load_json_file(args.json_file),
+            )
+        )
+        return
+
+    if args.command == "task":
+        print_json(
+            request_json(
+                MODULE_SPEC,
+                method="GET",
+                base_url=base_url,
+                api_key=api_key,
+                path=f"/api/v1/tasks/{args.task_id}",
+            )
+        )
+        return
+
+    if args.command == "results":
+        print_json(
+            request_json(
+                MODULE_SPEC,
+                method="GET",
+                base_url=base_url,
+                api_key=api_key,
+                path=f"/api/v1/tasks/{args.task_id}/results",
+            )
+        )
+        return
+
+    if args.command == "text":
+        task_response = request_json(
+            MODULE_SPEC,
+            method="GET",
+            base_url=base_url,
+            api_key=api_key,
+            path=f"/api/v1/tasks/{args.task_id}",
+        )
+        task_data = task_response.get("data", {}) if isinstance(task_response, dict) else {}
+        output = _normalize_output(task_data.get("output"))
+        json_url = output.get("json_url")
+        if not json_url:
+            raise SystemExit("当前 task 还没有可用的 json_url，请先确认任务已完成")
+
+        payload = _fetch_transcription_payload(str(json_url))
+        sentences = ((payload.get("timecodes") or {}).get("sentences") or [])
+        print_json(
+            {
+                "task_id": args.task_id,
+                "status": task_data.get("status"),
+                "json_url": json_url,
+                "text_field": "text",
+                "sentence_field": "timecodes.sentences[*].text",
+                "duration_ms": payload.get("duration_ms"),
+                "sentence_count": len(sentences) if isinstance(sentences, list) else None,
+                "text": payload.get("text", ""),
+            }
+        )
+        return
+
+
+if __name__ == "__main__":
+    main()