Skip to contents

Assumes fetch_issue_authors() has already been run, so each repo's creation timestamp can be read straight out of its issue-authors.csv rather than fetched again here. A repo missing from that file defaults back to date_start.

Usage

fetch_repo_commits(
  repo_urls,
  out_dir,
  batch_size = 50L,
  date_start = as.Date("2015-01-01"),
  date_end = NULL
)

Arguments

repo_urls

Character vector of repo URLs to fetch issue authors for. Assumed already ordered by priority if it matters which get fetched first (see interlace_for_even_coverage()).

out_dir

Directory to read/write the CSV + done-list checkpoint files.

batch_size

Repos fetched (concurrently) per checkpoint.

date_start, date_end

Passed to github_commit_counts_by_month().

Value

A tibble with columns repo_url, month, n_commits - the full accumulated result, including rows from any previous run(s).

Examples

repo_urls <- c (
    "https://github.com/ropensci/targets",
    "https://github.com/ropensci/drake"
)
if (FALSE) { # \dontrun{
commit_counts_tbl <- fetch_repo_commits (repo_urls, "path/to/repo-data-out")
} # }