GitHub Action执行R爬虫脚本报错：nothing to commit, working tree clean

问题背景 #

通过GitHub Action每小时运行R爬虫脚本，根据ASIN向量抓取亚马逊商品名称，生成带时间戳的CSV文件并提交到仓库。首次运行正常，但后续执行时出现nothing to commit, working tree clean错误，导致Action执行失败。

报错原因 #

• git的commit命令在没有文件变更时会返回非零退出码，而GitHub Action默认将非零码判定为执行失败
• 可能伴随的问题：R脚本生成的文件名包含特殊字符（如冒号、空格），或脚本运行异常未生成CSV文件

修复方案 #

1. 调整git提交逻辑，允许无变更时跳过提交 #

修改Action中的Commit files步骤，添加条件判断，仅在有文件变更时执行提交和推送，避免无变更时报错：

git config --local user.name github-actions
git config --local user.email "actions@github.com"
git add data/*
# 检查是否有暂存的变更
if git diff --cached --quiet; then
  echo "No new files to commit, skipping push"
else
  git commit -m "GH ACTION Autorun $(date)"
  git push origin main
fi

或者简化为允许commit命令失败时不终止Action：

git commit -am "GH ACTION Autorun $(date)" || true

2. 优化R脚本的文件名生成与目录检查 #

• 生成无特殊字符的时间戳，避免文件系统或git识别问题：

# 生成格式为YYYYMMDD_HHMMSS的时间戳
date <- format(Sys.time(), "%Y%m%d_%H%M%S")
# 确保data目录存在
if (!dir.exists("data")) dir.create("data")
# 导出CSV，禁用行名减少冗余
write.csv(product_name_df_delay_ua, file = paste0("data/Amazon_name_", date, ".csv"), row.names = FALSE)

• 移除脚本中的全局赋值final_df <<- as.data.frame(final)，改用局部变量避免潜在冲突：

name_checker_2 <- function(asin) {
  url_compositor <- paste0("https://www.amazon.it/dp/", asin)
  Sys.sleep(15) # 用R原生的Sys.sleep替代system('sleep 15')
  page <- read_html(url_compositor, user_agent = ua)
  name <- page %>%
    html_nodes(xpath = "//*[@id='productTitle']/text()") %>%
    html_text() %>%
    trimws()
  # 直接返回数据框，无需全局变量
  return(data.frame(asin = asin, asin_2 = unlist(asin), name = name))
}

3. 优化Action的R包安装步骤 #

将多个单独的安装命令合并为批量安装，提升效率：

R -e 'install.packages(c("tidyverse", "tibble", "openxlsx", "gdata", "lubridate", "rvest", "stringr", "dplyr", "purrr", "plyr"), dependencies = TRUE)'

同时使用最新版本的Action组件：

- uses: actions/checkout@v4
- uses: r-lib/actions/setup-r@v2

修改后的完整YAML配置 #

# Hourly scraping
name: amazonR

on:
  schedule:
    - cron: '0 * * * *'

jobs:
  autoscrape:
    runs-on: macos-latest
    steps:
    - uses: actions/checkout@v4
    - uses: r-lib/actions/setup-r@v2

    - name: Install packages
      run: |
        R -e 'install.packages(c("tidyverse", "tibble", "openxlsx", "gdata", "lubridate", "rvest", "stringr", "dplyr", "purrr", "plyr"), dependencies = TRUE)'
            
    - name: Scrape
      run: Rscript amazon_name_checker_simplest.R

    - name: Commit files
      run: |
        git config --local user.name github-actions
        git config --local user.email "actions@github.com"
        git add data/*
        if git diff --cached --quiet; then
          echo "No changes to commit"
        else
          git commit -m "GH ACTION Autorun $(date)"
          git push origin main
        fi
      env:
        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

内容的提问来源于stack exchange，提问作者Andrea