generate_sysartifacts_results

src.scrapers.generate_sysartifacts_results

Generate sysartifacts-compatible results.md files for USENIX conferences.

This script scrapes USENIX technical-sessions pages to collect paper titles, artifact evaluation badges, and paper PDF URLs, then produces results.md files in the format expected by sysartifacts.github.io.

Supported conferences include any that follow the standard USENIX technical-sessions layout (e.g. FAST, OSDI, ATC/NSDI).

Usage

Generate results for FAST 2024 and 2025

python generate_sysartifacts_results.py --conference fast --years 2024,2025 --output_dir ./out

Generate results for OSDI 2024

python generate_sysartifacts_results.py --conference osdi --years 2024 --output_dir ./out

Preview without writing (print to stdout)

python generate_sysartifacts_results.py --conference fast --years 2025 --dry-run

Custom directory prefix (e.g. "usenixatc" instead of "atc")

python generate_sysartifacts_results.py --conference atc --years 2024 --dir-prefix usenixatc

Requirements

pip install requests beautifulsoup4 pyyaml lxml

The generated results.md files can be dropped into the sysartifacts repo: _conferences//results.md

generate_results_md(conference, year, papers_with_badges)

Generate a sysartifacts-compatible results.md for a USENIX conference year.

Parameters:

Name	Type	Description	Default
conference		Conference short name (e.g. 'fast', 'osdi', 'atc')	required
year		Conference year (int)	required
papers_with_badges		List of dicts with keys 'title', 'badges', 'paper_url' (as returned by usenix_scrape.scrape_conference_year)	required

Returns:

Type	Description
	String content of the results.md file.

Source code in src/scrapers/generate_sysartifacts_results.py

def generate_results_md(conference, year, papers_with_badges):
    """
    Generate a sysartifacts-compatible results.md for a USENIX conference year.

    Args:
        conference: Conference short name (e.g. 'fast', 'osdi', 'atc')
        year: Conference year (int)
        papers_with_badges: List of dicts with keys 'title', 'badges', 'paper_url'
                            (as returned by usenix_scrape.scrape_conference_year)

    Returns:
        String content of the results.md file.
    """
    # Count badges
    n_available = sum(1 for p in papers_with_badges if "available" in p["badges"])
    n_functional = sum(1 for p in papers_with_badges if "functional" in p["badges"])
    n_reproduced = sum(1 for p in papers_with_badges if "reproduced" in p["badges"])

    # Build YAML front-matter artifacts list
    artifact_lines = []
    for p in papers_with_badges:
        # Escape YAML special characters in title
        title = p["title"].replace('"', '\\"')
        badges = ",".join(p["badges"])
        paper_url = p.get("paper_url", "")

        artifact_lines.append(f'  - title: "{title}"')
        if paper_url:
            artifact_lines.append(f'    paper_url: "{paper_url}"')
        artifact_lines.append(f'    badges: "{badges}"')
        artifact_lines.append("")

    artifacts_yaml = "\n".join(artifact_lines).rstrip()

    md = f"""---
title: Results
order: 50
available_img: "usenix_available.svg"
available_name: "Artifacts Available"
functional_img: "usenix_functional.svg"
functional_name: "Artifacts Evaluated - Functional"
reproduced_img: "usenix_reproduced.svg"
reproduced_name: "Results Reproduced"

artifacts:
{artifacts_yaml}
---

**Evaluation Results**:

* {n_available} Artifacts Available
* {n_functional} Artifacts Functional
* {n_reproduced} Results Reproduced

<table>
  <thead>
    <tr>
      <th>Paper title</th>
      <th>Avail.</th>
      <th>Funct.</th>
      <th>Repro.</th>
    </tr>
  </thead>
  <tbody>
  {{% for artifact in page.artifacts %}}
    <tr>
      <td>
        {{% if artifact.paper_url %}}
          <a href="{{{{artifact.paper_url}}}}" target="_blank">{{{{artifact.title}}}}</a>
        {{% else %}}
          {{{{ artifact.title }}}}
        {{% endif %}}
      </td>
      <td>
        {{% if artifact.badges contains "available" %}}
          <img src="{{{{ site.baseurl }}}}/images/{{{{ page.available_img }}}}" alt="{{{{ page.available_name }}}}" width="50px">
        {{% endif %}}
      </td>
      <td>
        {{% if artifact.badges contains "functional" %}}
          <img src="{{{{ site.baseurl }}}}/images/{{{{ page.functional_img }}}}" alt="{{{{ page.functional_name }}}}" width="50px">
        {{% endif %}}
      </td>
      <td>
        {{% if artifact.badges contains "reproduced" %}}
          <img src="{{{{ site.baseurl }}}}/images/{{{{ page.reproduced_img }}}}" alt="{{{{ page.reproduced_name }}}}" width="50px">
        {{% endif %}}
      </td>
    </tr>
  {{% endfor %}}
  </tbody>
</table>
"""
    return md

generate_organizers_md(organizers)

Generate a sysartifacts-compatible organizers.md from scraped organizer data.

Parameters:

Name	Type	Description	Default
organizers		Dict with 'chairs' and 'members' lists, each containing {'name': ..., 'affiliation': ...} dicts.	required

Returns:

Type	Description
	String content of the organizers.md file, or None if no data.

Source code in src/scrapers/generate_sysartifacts_results.py

def generate_organizers_md(organizers):
    """
    Generate a sysartifacts-compatible organizers.md from scraped organizer data.

    Args:
        organizers: Dict with 'chairs' and 'members' lists,
                    each containing {'name': ..., 'affiliation': ...} dicts.

    Returns:
        String content of the organizers.md file, or None if no data.
    """
    if not organizers or (not organizers.get("chairs") and not organizers.get("members")):
        return None

    lines = [
        "---",
        "title: Organizers",
        "order: 20",
        "---",
        "",
        "## Artifact Evaluation Committee Co-Chairs",
        "",
    ]

    for chair in organizers.get("chairs", []):
        aff = f", {chair['affiliation']}" if chair["affiliation"] else ""
        lines.append(f"{chair['name']}{aff} <br>")

    lines.append("")
    lines.append("## Artifact Evaluation Committee")
    lines.append("")

    members = organizers.get("members", [])
    for i, member in enumerate(members):
        aff = f", {member['affiliation']}" if member["affiliation"] else ""
        suffix = "<br>" if i < len(members) - 1 else ""
        lines.append(f"{member['name']}{aff}{suffix}")

    return "\n".join(lines) + "\n"