Cheatsheet: Zip

Pass one or more glob patterns after -x: zip -r archive_name.zip /path/to/directory -x '.log' '.tmp' 'node_modules/'. Each pattern matches against the path stored in the archive, so quote it to stop the shell from expanding it first, and use a trailing / to exclude everything under a directory while still creating the directory entry itself.

-0 stores files with no compression (fastest, largest output) and -9 uses maximum Deflate compression (slowest, smallest output); the default is -6, a balance of speed and size. Use -0 for files that are already compressed (video, images, other zip/gzip archives) since re-compressing them wastes CPU time for little or no size reduction, and often makes already-compressed data very slightly larger.

Use -j (junk paths) to store only the file names, discarding all directory components: zip -j archive_name.zip /path/to/directory/*. This is useful when you want every matched file dropped flat into the archive root, but be aware it will silently overwrite entries if two source files share the same base name.

Yes -- use -s with a size suffix: zip -s 100m -r archive_name.zip /path/to/directory splits the output into 100 MB volumes (archive_name.z01, .z02, ..., and a final archive_name.zip index). To recombine and extract, first run zip -s 0 archive_name.zip --out combined.zip to merge the volumes, then unzip combined.zip.

Use zip -sf archive_name.zip to print just the stored file list, or the more detailed zipinfo archive_name.zip (or unzip -l archive_name.zip) for per-entry size, compression ratio, and timestamp. Both read the archive's central directory only, so they work even on very large archives without extracting anything.