signfetch v0.2.0a1

Python package focused on discovering and downloading data via FAIR Signposting.

Scope

Version 0.2.0a1 exposes two public APIs:

• list_item_links(target) – discover item links without downloading
• download_data(target, output_dir=...) – discover and download data resources

Both functions use FAIR Signposting discovery mechanisms.

The package discovers item links from the three main Signposting delivery mechanisms:

1. HTTP Link headers
2. HTML <link> elements
3. Link Sets discoverable with the linkset relation

Link sets are automatically retrieved and parsed when present.

---

Example – discover data resources

You can inspect available data resources without downloading them.

from signfetch import list_item_links

links = list_item_links(
    "https://doi.org/10.5281/zenodo.12542566"
)

for link in links:
    print(link.url, link.source)

---

Example – download discovered data

download_data() performs discovery and downloads all discovered resources.

from pathlib import Path
from signfetch import download_data

result = download_data(
    "https://doi.org/10.5281/zenodo.12542566",
    output_dir=Path("downloads")
)

print(result.unique_item_count)

for item in result.items:
    print(item.url, item.saved_path)

---

Design notes

Main components:

Signposting discovery

• SignpostingHarvester – orchestrates discovery
• LinkHeaderParser – parses HTTP Link headers
• HtmlLinkParser – parses HTML <link> elements
• LinksetParser – parses linkset documents

Downloading

• DataDownloader – downloads resources discovered via item links
• downloads are executed concurrently

Public API

• list_item_links() – discovery only
• download_data() – discovery + download

---

Installation

pip install signfetch