refactor commands that accept multiple input files to use improved process_input helper

[jqnatividad]

As per the function doc of process_input

/// Process the input files and return a vector of paths to the input files
///
/// If the input is empty, try to copy stdin to a file named stdin in the passed temp directory
/// If the input is empty and stdin is empty, return an error
/// If it's not empty, check the input files if they exist, and return an error if they don't
///
/// If the input is a directory, add all the files in the directory to the input
/// If the input is a file with the extension ".infile-list" read the file, and add each line as a
/// file to the input If the input is a file, add the file to the input
/// If the input are snappy compressed files, uncompress them before adding them to the input

resolves #1293 for cat. headers command also now uses process_input and gained infile-list support along with other process_input capabilities.

Commands that previously used process_input now also support .infile-list files - sqlp and to commands.

[derekmahar]

The directory argument is a nice addition. Now commands can operate on all CSV files in a directory that contains a very large number of files instead of providing a text file that contains the names of these files. A directory is itself a file that contains a list of file names!