Metadata-Version: 2.4
Name: paperd
Version: 0.1.1
Summary: Daily arXiv source-package downloader with SQLite persistence and Feishu webhook reporting
Author-email: Phil Fan <hw.phil.fan@gmail.com>
Requires-Python: >=3.11
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: inquirerpy>=0.3.4
Requires-Dist: prompt-toolkit<4,>=3.0
Requires-Dist: tqdm>=4.66.0
Dynamic: license-file

# 📃 paperd

📃 CLI tool for downloading arXiv papers and preparing for research thinking, with Feishu Webhook notification.

- 按 **Asia/Shanghai** 日期抓取 arXiv 论文元数据
- 用 API 拉元数据后，按 id 从 `https://arxiv.org/src/<id>` 下载源码包（`tar.gz`）并解压
- 将元数据与本地源码路径持久化到 SQLite
- 通过飞书 Webhook 机器人进行通知

## Quick Start

```bash
mkdir -p "${XDG_CONFIG_HOME:-$HOME/.config}/.paperd"
```

```bash
uv sync
uv tool install .
paperd init
paperd run
paperd run --date 2026-04-01
paperd run --from 2026-04-01 --sep 2
```

- 飞书通知默认关闭，仅在 `paperd init` 中配置 `FEISHU_WEBHOOK_URL` 后启用
- `paperd init` 会同时交互写入 `.env`、`.area`，并询问 `DATA_ROOT`
- 默认配置目录：`${XDG_CONFIG_HOME:-$HOME/.config}/.paperd`
- 默认数据目录：`${XDG_DATA_HOME:-$HOME/.local/share}/paperd`

### 查询已抓取论文

```bash
uv run paperd list --limit 10
uv run paperd list --category cs.AI --date 2026-04-03
uv run paperd show 2504.12345v1
```

### Area 配置

详见 [arXiv 分类列表](https://arxiv.org/category_taxonomy)。

- AI 安全与对齐 (cs.AI, cs.CR, cs.CY, stat.ML, cs.LG)
- 多模态模型 / VLM / Vision-Language Alignment (cs.CV, cs.CL, cs.AI, cs.LG, cs.IR)
- 大语言模型推理加速 (cs.DC, cs.AR, cs.CE, eess.SP, cs.SY)
- VLA 安全与推理加速 (cs.AI, cs.CR, cs.CV, cs.CL, cs.DC, cs.CY)
- 云原生 AI / ML Systems / Serving (cs.DC, cs.CE, cs.SE, cs.SY, cs.NI, cs.CY)

## 开发测试

```bash
uv run python -m unittest discover -s tests -v
uv run python -m paperd --help
```
