I have slightly different parsers for each family of journals. I use the DOI to get the metadata where possible. Then I reformat to show title, authors and journal consistently. I also create a direct link to the PDF where possible because I prefer to get at the paper with a single click.