Data in Brief Material for Experimental Reproducibility

This dataset provides a comprehensive overview of software packages, their versions, and associated vulnerabilities across four major ecosystems: PyPI (Python), RubyGems (Ruby), Cargo Crates (Rust), and NPM (JavaScript). It includes detailed metadata for over 4,437,679 million unique packages and 60,950,846 versions and identifies vulnerable dependencies using 270,430 known vulnerabilities indexed in Open Source Vulnerabilities (OSV). The dataset covers both direct and transitive vulnerabilities, along with severity classifications. For each ecosystem, the versions of each package are flagged and annotated with vulnerability data, enabling risk analysis and supply chain security assessments. This resource supports research and tool development in software vulnerability management, dependency analysis, and security automation. The content of this folder is divided into raw and data folders 1. The raw folder contains a docker-compose.yml file that together with the command ‘docker compose up --build’ if run inside this folder, raises two containers. One with a MongoDB database and one with a Neo4J database, containing all the data extracted from the vunerabilities and package managers respectively. 2. The data folder contains all the accumulated content of the databases in a csv extension file. 3. In the file querys.cypher are shown Neo4J querys that can be used to extract information about packages and vulnerabilities.本数据集全面涵盖PyPI（Python）、RubyGems（Ruby）、Cargo Crates（Rust）及NPM（JavaScript）四大主流软件包生态系统中的软件包、版本及其关联漏洞信息。其包含超过4,437,679百万个唯一软件包与60,950,846个版本的详细元数据，并通过开源漏洞数据库（Open Source Vulnerabilities, OSV）收录的270,430个已知漏洞，识别出存在风险的依赖项。该数据集覆盖直接依赖与传递依赖漏洞，并附带严重性分级信息。针对每个生态系统，各软件包的版本均会被标记并附加漏洞数据，可用于开展风险分析与供应链安全评估。本资源可为软件漏洞管理、依赖项分析及安全自动化领域的研究与工具开发提供支撑。 本数据集所在文件夹分为raw与data两个子目录： 1. raw目录包含docker-compose.yml文件，在该目录内执行命令`docker compose up --build`即可启动两个容器：一个搭载MongoDB数据库，另一个搭载Neo4J数据库，分别存储从漏洞数据源与软件包管理器提取的全部数据。 2. data目录包含以CSV格式存储的数据库全量内容文件。 3. querys.cypher文件中收录了可用于提取软件包与漏洞相关信息的Neo4J查询语句。