The first input document serves as the basis for the combine operation. The content of every following input document is appended to the existing combined content, and the metadata of every following document replaces that of the previous documents (any metadata for which keys don't exist in the following documents is retained). A single output document with the combined content and metadata is output.