intellif-aihub 0.1.1__py3-none-any.whl → 0.1.2__py3-none-any.whl

aihub/services/tag_management.py

@@ -0,0 +1,35 @@
+# !/usr/bin/env python
+# -*-coding:utf-8 -*-
+from __future__ import annotations
+
+import httpx
+
+from ..exceptions import APIError
+from ..models.common import APIWrapper
+from ..models.tag_management import *
+
+_BASE = "/tag-resource-management/api/v1"
+
+
+class TagManagementService:
+    def __init__(self, http: httpx.Client):
+        self._project = _Project(http)
+
+    def select_projects(self) -> List[Project]:
+        return self._project.select_projects()
+
+    @property
+    def project(self) -> _Project:
+        return self._project
+
+
+class _Project:
+    def __init__(self, http: httpx.Client):
+        self._http = http
+
+    def select_projects(self) -> List[Project]:
+        resp = self._http.get(f"{_BASE}/select-projects")
+        wrapper = APIWrapper[ProjectListData].model_validate(resp.json())
+        if wrapper.code != 0:
+            raise APIError(f"backend code {wrapper.code}: {wrapper.msg}")
+        return wrapper.data.data

aihub/services/task_center.py

@@ -0,0 +1,190 @@
+from __future__ import annotations
+
+import datetime
+
+import httpx
+from loguru import logger
+
+from .tag_management import TagManagementService
+from .user import UserService
+from ..exceptions import APIError
+from ..models.common import APIWrapper
+from ..models.task_center import (
+    CreateTaskReq,
+    CreateTaskResp,
+    CreateTaskOtherInfo,
+    LabelProjectTypeEnum,
+    TaskCenterPriorityEnum,
+    LabelTaskDetail,
+)
+from ..models.user import UserSearchReq
+
+_BASE = "/task-center/api/v1"
+
+
+def date_str_to_timestamp(date_str: str) -> int:
+    date_obj = datetime.datetime.strptime(date_str, "%Y-%m-%d")
+    return int(date_obj.timestamp())
+
+
+class TaskCenterService:
+
+    def __init__(self, http: httpx.Client):
+        self._TaskCenter = _TaskCenter(http)
+        self._http = http
+
+    def create(self, payload: CreateTaskReq) -> int:
+        return self._TaskCenter.create(payload)
+
+    def get(self, task_id: int) -> LabelTaskDetail:
+        """获取任务信息
+
+        Args:
+            task_id (int): 任务ID
+
+
+        Returns:
+            LabelTaskDetail: 任务信息
+
+        """
+
+        return self._TaskCenter.get(task_id)
+
+    # 如果想要访问子对象，也保留属性
+    @property
+    def TaskCenter(self) -> _TaskCenter:
+        return self._TaskCenter
+
+    def create_label_task(
+        self,
+        name: str,
+        dataset_version_name: str,
+        feishu_doc_name: str,
+        task_receiver_name: str,
+        estimated_delivery_at: str,
+        project_name: str,
+        label_type: LabelProjectTypeEnum = LabelProjectTypeEnum.IMAGE_CLASSIFICATION,
+        description: str = "",
+        task_priority: TaskCenterPriorityEnum = TaskCenterPriorityEnum.low,
+    ) -> int:
+        """创建标注任务
+
+        Examples:
+            >>> from aihub.client import Client
+            >>> client = Client(base_url="xxx", token="xxxx")
+            >>> task_id = client.task_center.create_label_task( \
+name="test_task", dataset_version_name="re/V1",\
+feishu_doc_name="人脸质量人脸照片分类", task_receiver_name="hyc", \
+project_name="hycpro", label_type=LabelProjectTypeEnum.IMAGE_CLASSIFICATION, description="test_description", \
+task_priority="low", estimated_delivery_at= "2025-08-01")
+            1
+
+        Args:
+            name (str): 任务名称
+            dataset_version_name (str): 数据集版本名称
+            feishu_doc_name (str): 飞书文档名称
+            task_receiver_name (str): 任务接收者名称
+            estimated_delivery_at (str): 预计交付时间，格式为 "YYYY-MM-DD"
+            project_name (str): 项目名称
+            label_type (LabelProjectTypeEnum): 标注项目类型，默认为图像分类
+            description (str): 任务描述，默认为空
+            task_priority (TaskCenterPriorityEnum): 任务优先级，默认为低优先级
+        Returns:
+            任务ID
+        """
+        # 获取接收者ID
+        user_service = UserService(self._http)
+        task_receiver_id = user_service.search_one(
+            payload=UserSearchReq(nickname=task_receiver_name)
+        )
+
+        # 获取项目ID
+        tag_service = TagManagementService(self._http)
+        projects = tag_service.select_projects()
+        project_id = None
+        for project in projects:
+            if project.name == project_name:
+                project_id = project.id
+                break
+
+        if project_id is None:
+            raise APIError(f"未找到项目: {project_name}")
+
+        # 获取数据集ID
+        from .dataset_management import DatasetManagementService
+
+        dataset_service = DatasetManagementService(self._http)
+        dataset_version = dataset_service.get_dataset_version_by_name(
+            version_name=dataset_version_name
+        )
+
+        dataset_id = dataset_version.dataset_id
+        dataset_version_id = dataset_version.id
+
+        # 获取文档ID
+        from .document_center import DocumentCenterService
+
+        doc_service = DocumentCenterService(self._http)
+        docs = doc_service.get_documents(name=feishu_doc_name)
+
+        if not docs:
+            raise APIError(f"未找到文档: {feishu_doc_name}")
+
+        doc_id = docs[0].id
+
+        # 创建任务
+        other_info = CreateTaskOtherInfo(
+            label_project_type=label_type,
+            dataset_id=dataset_id,
+            dataset_version_id=dataset_version_id,
+            doc_id=doc_id,
+            doc_type="doc_center",
+        )
+        estimated_delivery_at_timestamp = date_str_to_timestamp(estimated_delivery_at)
+        task_req = CreateTaskReq(
+            name=name,
+            description=description,
+            task_priority=task_priority,
+            type="label",
+            receiver_id=task_receiver_id,
+            other_info=other_info,
+            project_id=project_id,
+            estimated_delivery_at=estimated_delivery_at_timestamp,
+        )
+
+        return self.create(task_req)
+
+
+class _TaskCenter:
+    def __init__(self, http: httpx.Client):
+        self._http = http
+
+    def create(self, payload: CreateTaskReq) -> int:
+        """创建任务"""
+        logger.debug(f"create task: {payload}")
+        resp = self._http.post(
+            f"{_BASE}/tasks",
+            json=payload.model_dump(),
+        )
+        logger.debug(f"create task response: {resp.text}")
+        if resp.status_code != 200:
+            raise APIError(
+                message="API Error", status=resp.status_code, detail=resp.json()
+            )
+        wrapper = APIWrapper[CreateTaskResp].model_validate(resp.json())
+
+        return wrapper.data.id
+
+    def get(self, task_id: int) -> LabelTaskDetail:
+        """获取任务"""
+        logger.debug(f"get task: {task_id}")
+        resp = self._http.get(
+            f"{_BASE}/tasks/{task_id}",
+        )
+        logger.debug(f"get task response: {resp.text}")
+        if resp.status_code != 200:
+            raise APIError(
+                message="API Error", status=resp.status_code, detail=resp.json()
+            )
+        wrapper = APIWrapper[LabelTaskDetail].model_validate(resp.json())
+        return wrapper.data

aihub/services/user.py

@@ -0,0 +1,47 @@
+# !/usr/bin/env python
+# -*-coding:utf-8 -*-
+from __future__ import annotations
+
+import httpx
+
+from ..exceptions import APIError
+from ..models.common import APIWrapper
+from ..models.user import *
+
+_BASE = "/api/v1/search-users"
+
+
+class UserService:
+
+    def __init__(self, http: httpx.Client):
+        self._user = _User(http)
+
+    # 直接把常用方法抛到一级，调用体验简单
+    def search_one(self, payload: UserSearchReq) -> int:
+        return self.user.search(payload)
+
+    # 如果想要访问子对象，也保留属性
+    @property
+    def user(self) -> _User:
+        return self._user
+
+
+class _User:
+    def __init__(self, http: httpx.Client):
+        self._http = http
+
+    def search(self, payload: UserSearchReq) -> int:
+        resp = self._http.post(
+            f"{_BASE}",
+            json=payload.model_dump(by_alias=True, exclude_none=True),
+        )
+        wrapper = APIWrapper[UserSearchListData].model_validate(resp.json())
+        if wrapper.code != 0:
+            raise APIError(f"backend code {wrapper.code}: {wrapper.msg}")
+        if wrapper.data.total == 0:
+            raise APIError("no dataset found")
+        for item in wrapper.data.data:
+            if item.nickname == payload.nickname:
+                return item.id
+        else:
+            raise APIError("no user found")

aihub/utils/download.py

@@ -0,0 +1,69 @@
+from __future__ import annotations
+
+import concurrent.futures
+import os
+import tempfile
+from typing import List, TypedDict
+
+import pyarrow.parquet as pq
+from tqdm import tqdm
+
+from .http import http_download_file
+from .s3 import s3_to_url
+
+
+class DatasetParquetMeta(TypedDict):
+    parent_dir: str
+    name: str
+    s3path: str
+    type: int  # 0=file, 1=dir
+
+
+_ENUM_FILE = 0
+
+
+def _read_parquet_index(file_path: str) -> List[DatasetParquetMeta]:
+    table = pq.read_table(file_path)
+    return table.to_pylist()  # 每行转 dict
+
+
+def _safe_rel(part: str) -> str:
+    if not part:
+        return ""
+    drive, tail = os.path.splitdrive(part)
+    return tail.lstrip(r"\/")
+
+
+def dataset_download(index_url: str, local_dir: str, worker: int = 4) -> None:
+    with tempfile.TemporaryDirectory() as tmpdir:
+        tmp_file = os.path.join(tmpdir, "index.parquet")
+        http_download_file(index_url, tmp_file)
+        rows = _read_parquet_index(tmp_file)
+
+    host = (index_url.split("//", 1)[-1]).split("/", 1)[0]
+
+    files = [
+        (
+            os.path.join(
+                local_dir,
+                _safe_rel(row["parent_dir"]),
+                _safe_rel(row["name"]),
+            ),
+            s3_to_url(row["s3path"], host),
+        )
+        for row in rows if row["type"] == _ENUM_FILE
+    ]
+
+    if worker < 1:
+        worker = 1
+
+    with tqdm(total=len(files), desc="Downloading dataset") as bar, \
+            concurrent.futures.ThreadPoolExecutor(max_workers=worker) as pool:
+
+        def _one(flocal: str, furl: str):
+            http_download_file(furl, flocal)
+            bar.update()
+
+        futures = [pool.submit(_one, p, u) for p, u in files]
+        for fut in concurrent.futures.as_completed(futures):
+            fut.result()

aihub/utils/http.py

@@ -0,0 +1,13 @@
+from __future__ import annotations
+
+import httpx
+import os
+
+
+def http_download_file(url: str, dst_path: str, chunk: int = 1 << 16) -> None:
+    os.makedirs(os.path.dirname(dst_path), exist_ok=True)
+    with httpx.stream("GET", url, follow_redirects=True, timeout=None) as r:
+        r.raise_for_status()
+        with open(dst_path, "wb") as f:
+            for block in r.iter_bytes(chunk):
+                f.write(block)

aihub/utils/s3.py

@@ -0,0 +1,77 @@
+from __future__ import annotations
+
+import os
+import re
+from pathlib import Path
+from typing import Any
+
+from loguru import logger
+from minio import Minio
+
+
+def s3_to_url(s3_path: str, host: str) -> str:
+    key = s3_path.replace("s3://", "").lstrip("/")
+    return f"http://{host.rstrip('/')}/{key}"
+
+
+def S3_path_to_info(s3_path) -> tuple[str | Any, str | Any] | None:
+    if not s3_path.startswith("s3://"):
+        return None
+
+    pattern = r"s3://(?P<bucket>\w+)/(?P<objectname>.+)"
+
+    match = re.match(pattern, s3_path)
+
+    if match:
+        bucket = match.group("bucket")
+        objectname = match.group("objectname")
+        return bucket, objectname
+    return None
+
+
+def local_path_to_s3_key(work_dir: str, local_path: str) -> str:
+    work_dir = Path(work_dir)
+    local_path = Path(local_path)
+    s3_key = str(local_path.relative_to(work_dir))
+    return s3_key
+
+
+def upload_dir_to_s3(
+    s3_client: Minio, local_dir: str, bucket: str, object_prefix: str
+) -> None:
+    logger.info(
+        f"Uploading directory {local_dir} to S3 bucket {bucket} with prefix {object_prefix}"
+    )
+
+    for root, dirs, files in os.walk(local_dir):
+        for file in files:
+            local_path = Path(root) / file
+            s3_key = local_path_to_s3_key(local_dir, str(local_path))
+            s3_client.fput_object(
+                bucket, os.path.join(object_prefix, s3_key), str(local_path)
+            )
+
+    logger.info(
+        f"Uploaded directory {local_dir} to S3 bucket {bucket} with prefix {object_prefix}"
+    )
+    return
+
+
+def download_dir_from_s3(
+    s3_client: Minio, bucket: str, object_prefix: str, local_dir: str
+) -> None:
+    logger.info(
+        f"Downloading directory from S3 bucket {bucket} with prefix {object_prefix} to {local_dir}"
+    )
+    objs = s3_client.list_objects(bucket, object_prefix, recursive=True)
+
+    for obj in objs:
+        file_name = Path(obj.object_name).relative_to(object_prefix)
+        s3_client.fget_object(
+            bucket, obj.object_name, os.path.join(local_dir, file_name)
+        )
+
+    logger.info(
+        f"Downloaded directory from S3 bucket {bucket} with prefix {object_prefix} to {local_dir}"
+    )
+    return

intellif_aihub-0.1.2.dist-info/METADATA

@@ -0,0 +1,110 @@
+Metadata-Version: 2.4
+Name: intellif-aihub
+Version: 0.1.2
+Summary: Intellif AI-hub SDK.
+Author-email: Platform Team <aihub@example.com>
+License-Expression: Apache-2.0
+Keywords: AI-hub,sdk,intellif
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: httpx<0.28,>=0.27
+Requires-Dist: pydantic<3.0,>=2.5.3
+Requires-Dist: typing-extensions<5.0,>=4.13.2
+Requires-Dist: pyarrow<16.0,>=15.0
+Requires-Dist: tqdm<5.0,>=4.66
+Requires-Dist: loguru>=0.7.3
+Requires-Dist: minio>=7.2.7
+Dynamic: license-file
+
+# Intellif AI-Hub SDK
+
+**Intellif AI-Hub** 官方 Python 开发包。  
+一个 `Client` 对象即可完成数据集管理、标注统计、任务中心等常见操作，无需手写 HTTP 请求。
+
+```
+aihub_sdk/
+├─ pyproject.toml
+├─ requirements.txt
+├─ src/aihub/
+│   ├─ client.py
+│   ├─ exceptions.py
+│   ├─ models/…
+│   ├─ services/…
+│   └─ utils/…
+└─ tests/
+```
+
+---
+
+## 💻 安装
+
+```bash
+# PyPI 安装
+pip install intellif-aihub
+# 运行环境：Python ≥ 3.8
+```
+
+---
+
+## 🚀 快速上手
+
+```python
+from aihub import Client
+
+BASE  = "http://192.168.13.160:30021"
+TOKEN = "eyJhb..."   # 或设置环境变量：export AI_HUB_TOKEN=...
+
+with Client(base_url=BASE, token=TOKEN) as cli:
+    # 1. 同时创建数据集 + 版本（上传本地 ZIP）
+    ds_id, ver_id, tag = cli.dataset_management.create_dataset_and_version(
+        dataset_name="cats",
+        is_local_upload=True,
+        local_file_path="/data/cats.zip",
+        version_description="first release",
+    )
+    print("数据集标识:", tag)  # 输出：cats/V1
+
+    # 2. 下载数据集
+    cli.dataset_management.run_download(
+        dataset_version_name=tag,
+        local_dir="/tmp/cats",
+        worker=8,
+    )
+
+    # 3. 获取标注平台全局统计
+    stats = cli.labelfree.get_project_global_stats("cat-project")
+    print("总标注数:", stats.global_stats.total_annotations)
+```
+
+---
+
+## 🌍 环境变量
+
+| 变量                       | 作用                                      | 默认值                           |
+|----------------------------|-------------------------------------------|----------------------------------|
+| `AI_HUB_TOKEN`             | API 鉴权 Token（可不在 `Client` 中显式传入） | –                                |
+
+---
+
+## 📦 打包 & 发布
+
+项目采用 PEP 517 / `pyproject.toml` 构建规范。
+
+```bash
+# 1️⃣ 构建 wheel / sdist
+python -m pip install --upgrade build
+python -m build                 # 生成 dist/*.whl dist/*.tar.gz
+
+# 2️⃣ 本地验证
+pip install --force-reinstall dist/*.whl
+python -c "import aihub, sys; print('SDK 版本:', aihub.__version__)"
+
+# 3️⃣ 发布到 PyPI 或私有仓库
+python -m pip install --upgrade twine
+twine upload dist/*
+```
+
+---

intellif_aihub-0.1.2.dist-info/RECORD

@@ -0,0 +1,32 @@
+aihub/__init__.py,sha256=YvuYzWnKtqBb-IqG8HAu-nhIYAsgj9Vmc_b9o7vO-js,22
+aihub/client.py,sha256=DJEG7fGAiWFuh-2-KucZdMxy2v3caDvCn-_1CuxZ90o,2953
+aihub/exceptions.py,sha256=l2cMAvipTqQOio3o11fXsCCSCevbuK4PTsxofkobFjk,500
+aihub/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+aihub/models/artifact.py,sha256=I07sXB3iIRlVnHm1HkYCXKIIPt5RGncxXuxANat8Vzs,3530
+aihub/models/common.py,sha256=qmabc2LkAdQJXIcpT1P35zxd0Lc8yDYdD4ame1iF4Bs,241
+aihub/models/dataset_management.py,sha256=lP92aOsZJihg4SEhf1jeITcXp-N8l_YzHYf1l9Zq7-g,3381
+aihub/models/document_center.py,sha256=xmAk_JIY3GjuVDZurMUonmSz3Siy3TAxhj3ewIJ6dUQ,489
+aihub/models/labelfree.py,sha256=hH9jOJ6_I3fdn1jlAtEuMwP3Vv9EzYlda_GKb8IJN2I,1179
+aihub/models/quota_schedule_management.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+aihub/models/tag_management.py,sha256=5_G9R6OZWYnm0ix_mLzMIujZhO_NNx6MeFA_Gl_zYBI,319
+aihub/models/task_center.py,sha256=N4Tdg_G3TgVc1Qse8sKbYihnbGkIVwHfGAEq-4v-zX4,3870
+aihub/models/user.py,sha256=Zv95o8ZAdAgmAsN5VsBhHRCCrw-RiAbHPg_Ljc-L3Mo,1004
+aihub/services/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+aihub/services/artifact.py,sha256=FaimaoHdZBr5hg6_jwj9md8EmxdWNFxIAFz1ML_xKu8,10370
+aihub/services/dataset_management.py,sha256=vNIc_JocgotFCFyFeCade0PWZspRjIkGhzt0qj1xQ0s,8672
+aihub/services/document_center.py,sha256=DClItxWXMMFPnKMR7kyNGohhXP3wowImj-Lm8vzBgNo,1339
+aihub/services/labelfree.py,sha256=Pc0kW-x8VLawqQWdY9ZDpZMPRB2KgXNk9-JE8Dz15Bs,1194
+aihub/services/quota_schedule_management.py,sha256=THSAJEi7QzilV9uT9Fy14p2ToIfn8nXQNgtIeEispvg,395
+aihub/services/reporter.py,sha256=ot93SmhxgwDJOzlHSCwlxDOuSydTWUEUQ-Ctp97wJBQ,669
+aihub/services/tag_management.py,sha256=Hci74medV5-p5VAZOYDEGrjSO1ioqET5IAnx7nqOiBo,943
+aihub/services/task_center.py,sha256=EA9j13IjQJYQMTNKDR-SbY-GUCfWoJmHc37ckVuPbUE,6110
+aihub/services/user.py,sha256=5qpvIKQhD0UWPQalbieLjyQY1cKp8pMlx_LTUKyIAKU,1327
+aihub/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+aihub/utils/download.py,sha256=Rh1m3VpMlw8-Kl36sowJ7M0dpB68u-9V4Vo3GQChq1I,1758
+aihub/utils/http.py,sha256=rSNh4uNP7E3YGm3H1indRHctxC5Wu5xNBPvDrb9UHt4,421
+aihub/utils/s3.py,sha256=ISIBP-XdBPkURpXnN56ZnIWokOOg2SRUh_qvxJk-G1Q,2187
+intellif_aihub-0.1.2.dist-info/licenses/LICENSE,sha256=QwcOLU5TJoTeUhuIXzhdCEEDDvorGiC6-3YTOl4TecE,11356
+intellif_aihub-0.1.2.dist-info/METADATA,sha256=hS0Y6s7g8IlicFlk7myET52gYJyFVRUGMo6FisCGqRY,2920
+intellif_aihub-0.1.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+intellif_aihub-0.1.2.dist-info/top_level.txt,sha256=vIvTtSIN73xv46BpYM-ctVGnyOiUQ9EWP_6ngvdIlvw,6
+intellif_aihub-0.1.2.dist-info/RECORD,,