pub fn read_dir(dir: &Path) -> Result<ApiData, ApiError>Expand description
Read every runs page and jobs?filter=all page in dir.
Names are sorted, then pages are merged. A later file that repeats a run
id is an error. A repeated job id counts once: the first record in file-name
order, then array order, is kept. Each job’s run_id must match the
jobs-<run_id> prefix of its file (jobs-15-2.json is run 15,
jobs-150.json is run 150). Files from attempts/{n}/jobs are not read.