Check the size of a github repo before downloading

check_github_repo_size

# http://webapps.stackexchange.com/questions/39587/view-estimated-size-of-github-repository-before-cloning
# tested on macOS
echo https://github.com/torvalds/linux.git | perl -ne 'print $1 if m!([^/]+/[^/]+?)(?:\.git)?$!' | xargs -I{} curl -s -k https://api.github.com/repos/'{}' | grep size

# output: 
# "size": 1746294,

curl -s https://api.github.com/repos/torvalds/linux | jq '.size' | numfmt --to=iec --from-unit=1024

What about private repos?

curl -s -H "Authorization: token GITHUB_TOKEN" https://api.github.com/repos/torvalds/linux | jq '.size' | numfmt --to=iec --from-unit=1024

curl -s -H "Authorization: Bearer <GITHUB_TOKEN>" https://api.github.com/repos/torvalds/linux | jq '.size' | numfmt --to=iec --from-unit=1024

I had to prefix Bearer last time I used this! 🎆

get_github_repo_size

#!/usr/bin/env -S bash -euo pipefail

get_github_repo_size() {

  if [[ "$1" == "--help" || "$1" == "-h" ]]; then
    echo "Usage: get_github_repo_size [OPTIONS] {GITHUB_REPO_URLS...}"
    echo
    echo "Retrieve and display the sizes of GitHub repositories."
    echo "The function will automatically use the GITHUB_TOKEN if it's set as an environment variable."
    echo
    echo "Warning: GitHub API repository size information may not be reliable."
    echo
    echo "Explanation:"
    echo
    echo "- GitHub support has indicated that due to Git Alternates and the way files are stored in GitHub repositories,"
    echo "  the numbers returned from the GitHub API cannot be relied upon for the actual size"
    echo "- Therefore, the size values returned by this script should not be considered"
    echo "  a reliable or accurate representation of the true repository size"
    echo "- The API size values may differ significantly from the actual repository size"
    echo
    echo "Options:"
    echo
    echo "  -h, --help     Display this help message."
    echo
    echo "Environment variables:"
    echo
    echo "  GITHUB_TOKEN   Use the provided GitHub token for authentication."
    echo
    echo "Exit codes:"
    echo
    echo "  0              Success"
    echo "  1              Fail"
    echo
    return
  fi

  local _i
  local  _err=0
  for _i in "${@}" ; do
    local _url="${_i}"
    local _owner="$(basename "$(dirname "${_url}")")"
    local _repo="$(basename "${_url}")"
    local _domain="$(basename "$(dirname "$(dirname "${_url}")")")"
    local _api_url="https://api.${_domain}/repos/${_owner}/${_repo}"

    local _headers=""
    if [ -n "$GITHUB_TOKEN" ]; then
      # Use the provided GitHub token if available
      local _headers="-H 'Authorization: Bearer $GITHUB_TOKEN'"
    fi

    # Fetch repository info, including size, in a single request
    local _response=$(curl -sS $_headers "$_api_url")

    # Check if the GitHub repo exists
    if jq -e '.message' <<< "$_response" >/dev/null; then
      # The API response contains a message field (indicating an error)
      local _message=$(jq -r '.message' <<< "$_response")
      >&2 printf 'error %s Failed to retrieve information about the GitHub repository: %s\n' "${_domain}/${_owner}/${_repo}" "${_message:-dns or http connection error}"
      _err=1
    else
      # The GitHub repo exists; fetch and format its size
      local _size=$(jq '.size' <<< "$_response")
      if [ "$_size" != "null" ]; then
        # Valid size found; format and print it
        printf '%s %s\n' "$(numfmt --to=iec --from-unit=1024 <<< "$_size")" "${_domain}/${_owner}/${_repo}"
      else
        >&2 printf 'error %s Invalid size data retrieved.\n'  "${_domain}/${_owner}/${_repo}"
        _err=1
      fi
    fi
  done
  >&2 echo
  >&2 echo "Warning: GitHub API repository size information may not be reliable (see --help for details)."
  return ${_err}
}

Example working usage:

$ get_github_repo_size "https://github.com/git/git" "https://github.com/danielmiessler/SecLists"

expected output:

To standard output Stream stdout only except Warning message to stderr

229M github.com/git/git
916M github.com/danielmiessler/SecLists

Warning: GitHub API repository size information may not be reliable (see --help for details)

Example failing usage:

$ get_github_repo_size "https://github.com/gitx/git" "https://github.com/git/gitx" "https://mygithub.net/git/git"

Expected output:

To standard error Stream stderr only

error github.com/gitx/git Failed to retrieve information about the GitHub repository: Not Found
error github.com/git/gitx Failed to retrieve information about the GitHub repository: Not Found
curl: (6) Could not resolve host: api.mygithub.net
error mygithub.net/git/git Failed to retrieve information about the GitHub repository: dns or http connection error

Warning: GitHub API repository size information may not be reliable (see --help for details)

I've developed a tool for checking GitHub repository sizes by leveraging the GitHub API.
You can find the tool here:
GitHub Repository Size Checker

Based on what I've read, GitHub support has indicated that due to Git Alternates and the way files are stored in GitHub repositories, the numbers returned from the GitHub API cannot be relied upon for the actual size.

I've developed a tool for checking GitHub repository sizes by leveraging the GitHub API. You can find the tool here: GitHub Repository Size Checker

Great tool!
Could you add checking based on the last commit?

I've developed a tool for checking GitHub repository sizes by leveraging the GitHub API. You can find the tool here: GitHub Repository Size Checker

I've been using this tool for quite some time, but it throws an error lately. So, I developed another web-based tool.
Site link: Repo Size Checker

I've developed a tool for checking GitHub repository sizes by leveraging the GitHub API. You can find the tool here: GitHub Repository Size Checker

I've been using this tool for quite some time, but it throws an error lately. So, I developed another web-based tool. Site link: Repo Size Checker

Thanks a lot for this!