Importing data from third-party version control systems

Importing projects from Mercurial

1. SSH into your GitHub Enterprise Server instance. For more information, see "Accessing the administrative shell (SSH)."

   $ ssh -p 122 admin@HOSTNAME

2. Make a raw clone of the project using the command below, specifying the URL of the source project, and a path to a temporary repository:

   $ git-import-hg-raw HG-CLONE-URL /PATH/REPO-NAME.git
   # Creates a new repository with one or more Git refs in "refs/import/" in the specified path.

3. Review the comma-separated (CSV) file in /PATH/REPO-NAME.git/git-import/raw-authors.csv. It should contain these columns:

   • ID: The author as stored in the original repository, followed by a unique identifier
   • NAME: The author as stored in the original repository

   To map authors from the original repository to an email address and name, create a new CSV file with the columns ID,(ignored),GIT_EMAIL,GIT_NAME, which replaces the author information for anything by "ID" with "GIT_EMAIL" and "GIT_NAME".

   Example:

   • Original author ID: octocat@111111-2222-3333-4444-55555555555

   • New email address: octocat@github.com

   • New name: The Octocat

     To map the original author to the new Git user, the CSV file should include the line:

     octocat@111111-2222-3333-4444-55555555555, ,octocat@github.com,The Octocat

4. Rewrite the authors and branches using the CSV file:

   $ git-import-rewrite --flavor hg --authors /PATH/AUTHORS-MAP-FILE.csv /PATH/REPO-NAME.git

5. If you haven't yet, create a new empty repository on GitHub Enterprise Server.

6. Change the current working directory to your local repository.

7. Push the imported repository to GitHub Enterprise Server:

   $ git push --mirror PUSH-URL-ON-GITHUB-ENTERPRISE

Importing projects from Subversion

1. SSH into your GitHub Enterprise Server instance. For more information, see "Accessing the administrative shell (SSH)."

   $ ssh -p 122 admin@HOSTNAME

2. Make a raw clone of the project using the command below, specifying the URL of the source project, and a path to a temporary repository:

   $ git-import-svn-raw SVN-CLONE-URL /PATH/REPO-NAME.git
   # Creates a new repository with one or more Git refs in "refs/import/" in the specified path.

3. Review the comma-separated (CSV) file in /PATH/REPO-NAME.git/git-import/raw-authors.csv. It should contain these columns:

   • ID: The author as stored in the original repository, followed by a unique identifier
   • NAME: The author as stored in the original repository

   To map authors from the original repository to an email address and name, create a new CSV file with the columns ID,(ignored),GIT_EMAIL,GIT_NAME, which replaces the author information for anything by "ID" with "GIT_EMAIL" and "GIT_NAME".

   Example:

   • Original author ID: octocat@111111-2222-3333-4444-55555555555

   • New email address: octocat@github.com

   • New name: The Octocat

     To map the original author to the new Git user, the CSV file should include the line:

     octocat@111111-2222-3333-4444-55555555555, ,octocat@github.com,The Octocat

4. Rewrite the authors and branches using the CSV file:

   $ git-import-rewrite --flavor svn --authors /PATH/AUTHORS-MAP-FILE.csv /PATH/REPO-NAME.git

5. If you haven't yet, create a new empty repository on GitHub Enterprise Server.

6. Change the current working directory to your local repository.

7. Push the imported repository to GitHub Enterprise Server:

   $ git push --mirror PUSH-URL-ON-GITHUB-ENTERPRISE

Importing projects from Team Foundation Version Control

1. SSH into your GitHub Enterprise Server instance. For more information, see "Accessing the administrative shell (SSH)."

   $ ssh -p 122 admin@HOSTNAME

2. Make a raw clone of the project using the command below, specifying the URL of the source project, and a path to a temporary repository:

   $ git-import-tfs-raw TEAM-FOUNDATION-CLONE-URL /PATH/REPO-NAME.git
   # Creates a new repository with one or more Git refs in "refs/import/" in the specified path.

3. Review the comma-separated (CSV) file in /PATH/REPO-NAME.git/git-import/raw-authors.csv. It should contain these columns:

   • ID: The author as stored in the original repository, followed by a unique identifier
   • NAME: The author as stored in the original repository

   To map authors from the original repository to an email address and name, create a new CSV file with the columns ID,(ignored),GIT_EMAIL,GIT_NAME, which replaces the author information for anything by "ID" with "GIT_EMAIL" and "GIT_NAME".

   Example:

   • Original author ID: octocat@111111-2222-3333-4444-55555555555

   • New email address: octocat@github.com

   • New name: The Octocat

     To map the original author to the new Git user, the CSV file should include the line:

     octocat@111111-2222-3333-4444-55555555555, ,octocat@github.com,The Octocat

4. Rewrite the authors and branches using the CSV file:

   $ git-import-rewrite --flavor tfs --authors /PATH/AUTHORS-MAP-FILE.csv /PATH/REPO-NAME.git

5. If you haven't yet, create a new empty repository on GitHub Enterprise Server.

6. Change the current working directory to your local repository.

7. Push the imported repository to GitHub Enterprise Server:

   $ git push --mirror PUSH-URL-ON-GITHUB-ENTERPRISE

Further reading

• "Command-line-utilities"