source-ingestion

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Source Ingestion

研究来源导入

Ingest sources so future analysis compounds instead of rediscovering the same document. The native source remains immutable; derived Markdown and wiki pages are working artifacts.

导入研究来源，以便后续分析可以基于已有成果展开，无需重复查找同一文档。原始来源文件保持不可更改；转换生成的Markdown和wiki页面为可编辑的工作产物。

Read First

必读文档

```
references/repository-contract.md
```
```
references/source-ledger.md
```
```
references/pdf-markdown-policy.md
```

```
references/repository-contract.md
```
```
references/source-ledger.md
```
```
references/pdf-markdown-policy.md
```

Workflow

工作流程

Identify source type, provenance, identifiers, and destination.
Store immutable originals under
```
sources/
```
,
```
data/raw/
```
,
```
data/external/
```
, or
```
reports/
```
.
If conversion is needed, hand off the conversion procedure to
```
document-conversion
```
; keep this workflow responsible for provenance and repository registration.
Add or update
```
sources/source-ledger.csv
```
.
Create or update
```
wiki/sources/<source_id>.md
```
.
Extract claims, methods, datasets, limitations, and open questions.
Update
```
wiki/index.md
```
,
```
wiki/log.md
```
, and any affected concept, claim, method, or question pages.

确定来源类型、出处、标识符及存储目标。
将不可更改的原始文件存储在
```
sources/
```
、
```
data/raw/
```
、
```
data/external/
```
或
```
reports/
```
目录下。
若需要转换，将转换流程移交至
```
document-conversion
```
；本工作流程负责记录来源出处及仓库注册信息。
添加或更新
```
sources/source-ledger.csv
```
文件。
创建或更新
```
wiki/sources/<source_id>.md
```
页面。
提取研究主张、方法、数据集、局限性及待解决问题。
更新
```
wiki/index.md
```
、
```
wiki/log.md
```
以及所有受影响的概念、主张、方法或问题页面。

Source ID

来源标识符（Source ID）

Prefer stable slugs:

```
smith-2024-topic
```
```
arxiv-2401-12345
```
```
doi-10-1145-short-topic
```
```
proposal-short-topic-2026
```

优先使用稳定的短标识：

```
smith-2024-topic
```
```
arxiv-2401-12345
```
```
doi-10-1145-short-topic
```
```
proposal-short-topic-2026
```

Conversion Handoff

转换移交

Preserve the native PDF.
Save derived Markdown under
```
sources/markdown/<source_id>.md
```
only after the conversion workflow records quality and command details.
Verify exact claims, tables, equations, and quotes against the native file.
If extraction quality is poor, mark it and avoid using the Markdown for exact evidence.

保留原始PDF文件。
仅在转换工作流程记录质量及命令详情后，将转换生成的Markdown文件保存至
```
sources/markdown/<source_id>.md
```
。
对照原始文件验证研究主张、表格、公式及引用内容的准确性。
若提取质量不佳，需标记该文件，且避免将其Markdown版本用作精确证据。

Output Contract

输出规范

The task is not complete until these are updated when applicable:

source file path
derived Markdown path
metadata or BibTeX path
```
sources/source-ledger.csv
```
```
wiki/sources/<source_id>.md
```
```
wiki/index.md
```
```
wiki/log.md
```

在适用情况下，需完成以下内容的更新，任务才算完成：

源文件路径
转换生成的Markdown文件路径
元数据或BibTeX文件路径
```
sources/source-ledger.csv
```
```
wiki/sources/<source_id>.md
```
```
wiki/index.md
```
```
wiki/log.md
```