CSBench

# Dataset Structure This is the dataset for detailed information of CSBench cases. Please refer to the CSBench Repo for running the experiments. An example row contains: - `id`: unique case identifier - `category`: top-level category of the case, one of ['AI', 'System', 'Application', 'Others'] - `sub_category`: finer-grained topic. - `university`: institution associated with the course - `course_id`: course identifier - `course_name`: course title - `programming_languages`: programming language list stored as a string - `lab_url`: task or lab description URL - `lab_name`: lab name - `course_url`: course homepage URL - `language`: normalized language label, either `English` or `Chinese` - `wiki_url`: related wiki reference URL - `pass_to_pass`: number of tests expected to keep passing - `fail_to_pass`: number of tests expected to be fixed - `task_desc`: concise natural-language description of the core task - `specification`: functionality requirements summarized from course materials - `scaffold_desc`: description of the provided source scaffold or implementation structure - `ut_desc`: summary of unit tests and the functionality they cover