first version for test

2 files changed, 116 insertions, 0 deletions

diff --git a/.github/workflows/daily_papers.yml b/.github/workflows/daily_papers.yml
index e69de29..d32db15 100644
--- a/.github/workflows/daily_papers.yml
+++ b/.github/workflows/daily_papers.yml
@@ -0,0 +1,27 @@
+name: Daily Paper Fetch
+
+on:
+  schedule:
+    - cron: '0 12 * * *'  # 每天UTC 12:00 触发
+  workflow_dispatch:       # 支持手动触发
+
+jobs:
+  fetch-and-update-readme:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: 3.10
+
+    - name: Install dependencies
+      run: |
+        pip install requests sentence-transformers feedparser PyGithub
+
+    - name: Run script
+      env:
+        TARGET_REPO_TOKEN: ${{ secrets.TARGET_REPO_TOKEN }}  # 之后我们会在主仓库配置这个
+        TARGET_REPO_NAME: "YourUsername/llm-bias-papers"     # 后面我们会改成你的目标仓库名
+      run: python scripts/fetch_papers.py
diff --git a/scripts/fetch_papers.py b/scripts/fetch_papers.py
index e69de29..49c3ba8 100644
--- a/scripts/fetch_papers.py
+++ b/scripts/fetch_papers.py
@@ -0,0 +1,89 @@
+import os
+import requests
+import feedparser
+import datetime
+from github import Github
+
+ARXIV_CATEGORIES = "cs.CL OR cs.AI OR stat.ML"
+KEYWORDS = ["LLM bias", "debias", "fairness", "equity", "inclusivity", "diversity", "ethical AI", "responsible AI"]
+
+TARGET_REPO_TOKEN = os.getenv("TARGET_REPO_TOKEN")
+TARGET_REPO_NAME = os.getenv("TARGET_REPO_NAME")
+
+def fetch_arxiv_papers():
+    """
+    从arXiv获取过去24小时的新论文
+    """
+    base_url = "http://export.arxiv.org/api/query"
+    # 使用带时区的UTC时间
+    now_utc = datetime.datetime.now(datetime.timezone.utc)
+    yesterday_utc = now_utc - datetime.timedelta(days=1)
+
+    params = {
+        "search_query": f"cat:{ARXIV_CATEGORIES}",
+        "sortBy": "submittedDate",
+        "sortOrder": "descending",
+        "max_results": 100
+    }
+    r = requests.get(base_url, params=params)
+    feed = feedparser.parse(r.content)
+
+    papers = []
+    for entry in feed.entries:
+        # entry.published 形如 "2025-03-28T10:05:24Z"
+        # 先解析出一个 naive datetime
+        published_naive = datetime.datetime.strptime(entry.published, "%Y-%m-%dT%H:%M:%SZ")
+        # 再为它添加UTC时区，使其成为 aware datetime
+        published_utc = published_naive.replace(tzinfo=datetime.timezone.utc)
+
+        if published_utc > yesterday_utc:
+            papers.append({
+                "title": entry.title,
+                "url": entry.link,
+                "abstract": entry.summary
+            })
+    return papers
+
+def filter_papers(papers):
+    relevant = []
+    for p in papers:
+        abstract_lower = p["abstract"].lower()
+        title_lower = p["title"].lower()
+        if any(kw.lower() in abstract_lower or kw.lower() in title_lower for kw in KEYWORDS):
+            relevant.append(p)
+    return relevant
+
+def update_readme_in_target(relevant_papers):
+    if not relevant_papers:
+        print("No relevant papers found. Skipping README update.")
+        return
+
+    g = Github(TARGET_REPO_TOKEN)
+    repo = g.get_repo(TARGET_REPO_NAME)
+
+    readme_file = repo.get_contents("README.md", ref="main")
+    readme_content = readme_file.decoded_content.decode("utf-8")
+
+    # 此处同样可以用带时区的时间来记录日期（格式化字符串不影响）
+    date_str = datetime.datetime.now(datetime.timezone.utc).strftime("%Y-%m-%d")
+    new_section = f"\n\n### {date_str}\n"
+    for p in relevant_papers:
+        new_section += f"- **[{p['title']}]({p['url']})**\n"
+
+    updated_content = readme_content + new_section
+
+    repo.update_file(
+        path="README.md",
+        message=f"Auto Update README with {len(relevant_papers)} papers ({date_str})",
+        content=updated_content,
+        sha=readme_file.sha,
+        branch="main"
+    )
+
+def main():
+    papers = fetch_arxiv_papers()
+    relevant_papers = filter_papers(papers)
+    update_readme_in_target(relevant_papers)
+
+if __name__ == "__main__":
+    main()