The Project Gutenberg recently put a new directory system in place. Now it is possible to reference their books easily and I propose to go for it. As for ISBN invent a "tag" (proposal: PGNO = Project Gutenberg No) and teach the parser to decode the number. Eplanation from the GUTINDEX.ALL file: The path is also based on the etext number, and is made up of single digits corresponding to all but the last digit in the filename. So, if you know the number, you know exactly where to find the eBook! The path for our example, 10010.txt, would be /1/0/0/1/10010/. Of course, the address of the download site would have to be prepended to the path; if the main PG site was being used, the full URL, including the filename, would be: http://ibiblio.org/gutenberg/1/0/0/1/10010/10010.txt The zip file of the above would be: http://ibiblio.org/gutenberg/1/0/0/1/10010/10010.zip Using the ftp site at iBiblio to obtain the zip: ftp://ibiblio.org/pub/docs/books/gutenberg/1/0/0/1/10010/10010.zip And you can see all the available files for #10010 at http://ibiblio.org/gutenberg/1/0/0/1/10010 As you can see, the first 4 digits are used as (sub)directory names; then the entire number is used as the final directory where all the formats of the book are to be found; convert "PGNO 10010" to: <a href="http://ibiblio.org/gutenberg/1/0/0/1/10010/">Project Gutenberg No 10010</a> -- | ,__o | _-\_<, http://www.gnu.franken.de/ke/ | (*)/'(*)