Many SVN repos have a trunk/tags/branches/ directory structure, so a code snapshot of the whole repository has a lot of duplicated files and takes lot longer to build. Download snapshot links for SVN should be aware of this structure (which can occur at the top level, or nested down a level or more e.g. https://sourceforge.net/p/funambol/code/HEAD/tree/ox-connector/tags/) and only generate a snapshot of the specific branch, tag, or the trunk that the current page is part of. If the current page is not within a branches/ tags/ or trunk/ URL, then use the top-level trunk/ dir if it exists and if it doesn't, then use the whole repo. I think that logic will work well, of course adjust it if needed.
The resulting zip files will have to incorporate the repo path in their filenames, so that there can be saved zip files of the same SVN revision but different directories.