checkin traverses the tree rooted at the current directory checking all files into the vendor branch of the repository directory located at CVSROOT/repository. If SCCS or RCS directories exist under the current directory, checkin checks out all files before committing them to the repository. Any RCS or SCCS directories are then ignored when the files are checked in.
checkin logs all updates to locally modified files, new files, and new directories to the standard output. If the -v option is specified it will log all updates instead of just the interesting ones.
When invoked with the -m flag, checkin takes the next argument to be the log message to use when checking in the files. The -f flag specifies the name of a file to use as the log message. If neither -m or -f is specified, checkin invokes your favorite editor for you to create a log message.
Each file is checked into the vendor branch specified by the vendor_tag on the command line. If the specified branch does not exist, checkin creates the branch 1.1.1 and labels it with the specified tag. As each file is checked in, it is labeled with the vendor_release_tag and logged with the specified log message.
checkin also understands about the way cvs removes files from a repository by moving them to an attic sub-directory. If a new version of a removed file is encountered, checkin will update the copy of the file in the attic and will log the update to the standard output. The file can be brought back into existence with the cvs add command (see the cvs man page for more details).
If checkin encounters a file with an extension that rcs(l) doesn't recognize that contains the rcs(l) $log$ header, it will prompt the terminal for a comment leader for that type of file.
Should know how to parse the CVSROOT/CVSROOT.adm/loginfo file to log all updates according to the same scheme as cvs itself.
Its a real CPU pig, but it works okay and isn't used nearly as much as other portions of cvs.