Feat(ovpack): recursively vectorize all imported docs

[sponge225]

Description

修复/添加 ovpack 导入后的“直接向量化入队”逻辑：导入完成后递归遍历整棵目录树，把子目录和内部文件都加入向量化入队，避免导入后在子目录范围内检索不到内容。同时使用固定 worker 池，降低大pack场景的协程创建与调度压力。

Related Issue

现状： import_ovpack(..., vectorize=True) 会触发“导入后向量化入队”，但原逻辑只对导入根目录做一次入队，导致：

• 限定 --uri=viking://resources/<import_root>/ 时检索经常为空
• 导入后的子目录/内部文件无法被召回（因为没有向量索引）

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional changes)
• Performance improvement
• Test update

Changes Made

• ovpack 导入 import_ovpack(..., vectorize=True) 后的向量化入队从“仅入队根目录”改为“递归覆盖导入树（目录 + 文件）”，导入后子目录检索可命中。
• 向量化标准对齐 add-resource：
  • 目录：读取并向量化该目录下存在的 .abstract.md / .overview.md （通过 vectorize_directory_meta 入队）。
  • 文件：通过 vectorize_file 对文件内容向量化（包含 max_input_chars 截断/格式处理等既有逻辑）。
• 并发模型优化：用 asyncio.Queue + 固定 worker_count 处理入队任务，避免一次性 gather(N) 创建海量 coroutine（对大 ovpack 更稳定）。

Testing

• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have tested this on the following platforms:
  • Linux
  • macOS
  • Windows

手动验证：

• 导入一个包含多级目录的 ovpack（ vectorize=True ），导入完成后执行：
• ov search --uri=viking://resources/<import_root>/ -n 10 "<关键词>"
• 预期：修改前子目录命中为空/很少；修改后子目录能命中（目录 .abstract/.overview 以及内部文件内容）。

提取root目录后再导入

修改前导入到了imports，修改后导入到了imports_fix

验证修改前

验证修改后

Checklist

• My code follows the project's coding style
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Screenshots (if applicable)

Additional Notes

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[github-actions]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🏅 Score: 85

🧪 No relevant tests

🔒 No security concerns identified

✅ No TODO sections

🔀 No multiple PR themes

⚡ Recommended focus areas for review Silent Exception Swallowing The code uses a bare except Exception: that returns without logging, which can hide real errors when reading directory metadata files (.abstract.md/.overview.md). except Exception: return

[github-actions]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
General	Add logging for directory metadata read errors Add structured logging when an exception occurs while reading directory metadata files (.abstract.md and .overview.md). Swallowing exceptions without logging makes debugging difficult. Use the existing logger to log the error with context (dir_uri). openviking/storage/local_fs.py [116-124] try: if await viking_fs.exists(abstract_uri, ctx=ctx): content = await viking_fs.read_file(abstract_uri, ctx=ctx) abstract = content.decode("utf-8") if isinstance(content, bytes) else content if await viking_fs.exists(overview_uri, ctx=ctx): content = await viking_fs.read_file(overview_uri, ctx=ctx) overview = content.decode("utf-8") if isinstance(content, bytes) else content -except Exception: +except Exception as e: + logger.warning("Failed to read directory metadata files", dir_uri=dir_uri, error=e) return Suggestion importance[1-10]: 6 __ Why: Adds structured logging for exceptions when reading directory metadata files, improving debuggability without altering core functionality.	Low
	Add worker error handling with logging Add error handling and logging for worker tasks that fail while indexing files or directories. This ensures we track individual task failures without crashing all workers, improving resilience and observability. openviking/storage/local_fs.py [142-155] async def worker() -> None: while True: kind, payload = await work_queue.get() try: if kind == "stop": return if kind == "dir": (dir_uri,) = payload await index_dir(dir_uri) elif kind == "file": file_uri, parent_uri, file_name = payload await index_file(file_uri, parent_uri, file_name) + except Exception as e: + logger.warning("Worker task failed", kind=kind, payload=payload, error=e) finally: work_queue.task_done() Suggestion importance[1-10]: 6 __ Why: Adds error handling and logging for worker tasks, enhancing resilience and observability of the indexing process.	Low

[qin-ctx]

这里为什么要加各种参数

[sponge225]

默认的node_limit是1000，level_limit是3，比较小，调大防止遍历不全

[kaisongli]

https://github.com/volcengine/OpenViking/actions/runs/24125971068/job/70390515057

curl -X GET -H "User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36" -H "Accept-Encoding: gzip, deflate" -H "Accept: */*" -H "Connection: keep-alive" -H "Accept-Language: zh-CN,zh;q=0.9" -H "X-OpenViking-Account: default" -H "X-OpenViking-User: default" -H "Authorization: [REDACTED]" -H "Content-Type: application/json" "http://127.0.0.1:1933/api/v1/fs/ls?uri=viking%3A%2F%2Fresources%2Fupload_2488f2b1a9894a3bb84ee425ae5ca9f9&simple=False&recursive=False"
windows系统运行有API报错，是否会影响这个接口的正常运行